Loading ...
Sorry, an error occurred while loading the content.
 

Re: [dubailug] Sata Drives

Expand Messages
  • Manu Abraham
    ... Go for the latest FC3. The newer 2.6.x kernel really do give a general performance boost.. ... Fast HDD are advised if you intend to do SQUID Object
    Message 1 of 29 , Jan 1, 2005
      On Sat January 1 2005 9:25 am, Rajeev wrote:
      > Manu/ cGull
      >
      > I will be purchasing the hardware this week. Could you please suggest me a
      > guideline spec if I use this machine for the following.
      >
      > Fedora C-2 or RH 9
      Go for the latest FC3. The newer 2.6.x kernel really do give a general
      performance boost..

      >
      > It works mainly as storage server. (SAMBA) (Backup to the main server -
      > i.e., every night the main server will copy the data to this server and no
      > user will be directly accessing this one - and it should be maximum cost
      > effective)
      >
      > It works also as Squid Proxy server for below 100 users.
      >
      Fast HDD are advised if you intend to do SQUID Object Caching .. In that case
      i would advise U320 SCA HDD's depending upon the users usage pattern

      For a proxy, having the Object cache on a RAID array/disk is meaningless.
      There is too much overhead. If you are looking at putting the Object cache on
      a RAID array/disk (if you have a dedicated controller, you can consider it as
      a RAID disk, rather than an array.. You have only /dev/sdxx, instead..), i
      would suggest using RAID 1(Mirrorring and Duplexing) as this will have a
      larger performance, since the writes are alternate. RAID 5 gives you only
      data transfer rate same as single disk
      whereas RAID 1 will give you very high availability.(If you are looking at
      RAID 1, i would strongly recomend a Hardware implementation.)

      Moreover, with a RAID 5 array, rebuilding would be more tedious in the case of
      a disk failure

      In this case i would suggest moving onto Hardware RAID solutions rather than a
      software RAID solution. The Software RAID solution makes use of the CPU's
      free processor cycles for replication.

      Whereas in a Hardware RAID solution, there is a dedicated processor and
      associated memory on the RAID board to achieve that.

      Maybe if you are cash strapped, you can look out for Zero channel RAID
      solutions as well which are still cost effective but still better than
      Software RAID.


      If that is the case i can suggest you some ASUS SMP server board with a zero
      channel/Full blown RAID, if you plan to build your own server. These boards
      are not very expensive and are quite reliable.. 5 years back, i used have
      have 2 machines, with dual PIII Xeon 800 Mhz with 512MB ECC SDRAM, with an
      Adaptec I20 Hardware RAID controller with U160 disk's, running RH6.0 and
      patches having about 120 users, and 146 PIII 800 MHz Machines as PVM cluster
      nodes and 17 SGI Visualization Workstations, for automobile modelling
      (CAE/CAM) purposes. That system is still running _unattended_, the reason
      there is a huge 40KVA UPS system with a battery room, that can last for 2
      days. I still give the guys a few pointers over phone ... The only trouble
      the guys compain is when the leased lines go down, when somebody breaks the
      roads..

      This machine was used as a mail server, SMB/NFS File server, Doubled as a
      DHCP/DNS, and a PVM cluster server ..


      > This will be a machine as you described - no monitor, no keyboard and no
      > mouse. I will have 1 GB of ram. And I need 1 terra byte space in Raid 5.
      > Since this machine will be part of a production network I can't do too many

      Put OS onto separate HDD's itself, if you are planning for extreme low
      downtime

      And user data into the RAID disk's/array.

      That kind of combinations depends upon what level of HA you are looking at

      For example, for my HA applications, i have a dedicated Hardware RAID U320
      controller with 2 raid arrays A & B..

      In my case it is not the number of users, but the bandwidth that is
      generated .. I have around 25 users only on that HA system as opposed to 100
      in your case.. And lot of financial data also....... That makes it HA..

      Array A contains OS and
      Array B contains User data..


      Another major suggestion that i can give you on your N/W design is that most
      probably you will not require to have a proxy...

      You can get around using Netfilter and NAT.. That will bring down a lot of
      your requirements... Then that machine would be a firewall machine doing NAT
      and PAT, but again it is not advisable to use a firewalling machine as your
      intranet server.

      Well, there are lot of points to be considered, like what exactly you are
      trying to achieve, the permutations and combinations are huge ..

      Manu
      > experiments.
      >
      > I would really appreciate your help,
      >
      >
      >
      >
      > -----Original Message-----
      > From: Manu Abraham [mailto:manu@...]
      > Sent: 31 December 2004 22:53
      > To: dubailug@yahoogroups.com
      > Subject: Re: [dubailug] Sata Drives
      >
      > On Fri December 31 2004 10:14 pm, cgull@... wrote:
      > > Manu,
      > >
      > > Thanks for the cool info!
      > > Can you tell us if it is possible to load this support via a module (for
      >
      > It depends upon the kernel .. I had SATA support into 2.4 also One of my
      > 2.4 machines which i use as my NFS/SAMBA fileserver, has around a total of
      > > 200Gigs n/w storage on SATA, i have another IBM one with U320 SCA SCSI
      > also.
      >
      > No problems also .. If somebody requires i can compile the kernel modules
      > if i
      > have that kernel version. But rather than picking up my binary modules i
      > would really suggest guys to put in the additional pain of recompiling.. It
      > is really worth it ..
      >
      > > example via insmod) rather than recompile kernel with changed options?
      >
      > Once compiled libata can be loaded as a module as i have loaded..
      > You need to have to enable SCSI support also to have SATA functioning ..
      >
      >
      > appletalk 33972 2
      > ax25 58476 2
      > ipx 26540 2
      > ipv6 237440 18
      > intel_mch_agp 10128 1
      > agpgart 30888 2 intel_mch_agp
      > bluetooth 45924 6 l2cap,hci_vhci,hci_uart,hci_usb
      > af_packet 20872 2
      > 3c59x 36904 0
      > binfmt_misc 10120 1
      > ata_piix 7684 0
      > libata 38016 1 ata_piix,[permanent]
      > sd_mod 20224 0
      > scsi_mod 108876 5 sg,st,sr_mod,libata,sd_mod
      >
      >
      > For older kernels, for libata support the __only__ way is to patch the
      > kernel
      > with the libata patch. If somebody wants to find the patch, it is
      > maintained by Jeff Garzick. You can find the patches on kernel.org. It
      > depends how old or what kernel it is to know what other patches if any
      > libata requires.. I had 2.4.24 or 25... I think i had a lot of patches onto
      > that..
      >
      > I had the ck-xx patches, libata patch, lot of patches, i don't recollect..
      > what all.. That machine is in a corner humming away without any problems at
      > all as i said for more than 8 months without any interruptions.. Even
      > during power up and downs. There is no keyboard/monitor/mouse on that..
      > Anything goes wrong (nothing has gone wrong except power supply blew up
      > once) somebody
      > just presses the reset button (in emergency) else restarts through SSH.. As
      > simple as that..
      >
      > I have a quite heavy data transfer on that machine GB's daily as machine
      > does
      > storage for some PVR jobs also.
      >
      > Regarding new kernels... There is no problem at all, It is like fish
      > jumping to water.
      >
      > The most problematic one was the ICH5, ie, SATA on the Intel boards, but
      > even
      > that, the problems are no longer there with the new kernels..
      >
      > Kernel-2.6.9 onwards beautiful support is there, but there is exception of
      > Seagate HDD's on SiL controllers, but only a firmware problem on the drives
      > as i said earlier ..
      >
      > Manu
      >
      > > cGull.
      > >
      > > Manu Abraham <manu@...> wrote on 31.12.2004, 18:49:14:
      > > > On Fri December 31 2004 7:30 pm, cgull@... wrote:
      > > > > Rajeev,
      > > > >
      > > > > I have tried SATA drives on HP hardware and I had some problems.
      > > >
      > > > Which controller was this.. It is known that Seagate HDD's and SiL
      > > > Controllers don't get along well, but therre have been reports that
      >
      > HDD's
      >
      > > > with newer firmware does not have any problems ..
      > > >
      > > > > For example Knoppix and Tomsrtbt both hung on boot because of the
      >
      > SATA.
      >
      > > > libata i believe .. Try using the SATA drives in compatibity mode,
      > > > eventhough you loose performance, but after installation of the OS,
      > > > just rebuild the kernel with the relevant patches..
      > > >
      > > > > My Mandrake distro has no problems with it however.
      > > >
      > > > Mandrake has libata support . :-)
      > > >
      > > > > I searched the internet and found there were some issues with SATA.
      > > > > To avoid this I just went for SCSI which is guaranteed to work with
      > > > > Linux.
      > > >
      > > > SATA also uses the same SCSI commands under linux... In fact for a SATA
      > > > HDD you have the regular /dev/sdxx for SATA as compared to SCSI>
      > > >
      > > > > I know SATA is cheaper, but then I wasnt the one forking out the cash
      > > > > so I could make the choice of SCSI instead.
      > > > > You may also have other issues such as SATA not being hot-pluggable
      > > > > from
      > > >
      > > > For SCSI also, unless you have SCA SCSI you cannot Hot swap... SCA SCSI
      > > > is still more expensive than standard SCSI, and moreover for SCA SCSI
      >
      > you
      >
      > > > have to have the drive cage to properly Hotswap.. The only difference
      > > > between SCA SCSI and SCSI is that the power to the SCSI disk comes
      > > > through the SCA SCSI connector rather than the power connector.. SCA
      >
      > SCSI
      >
      > > > has no separate power connectors, but they have a controller situated
      > > > in the drive cage to shutdown the power to the relevant drive ...
      > > >
      > > > > your server or disk array.
      > > > > Tell us what hardware you plan to use (server and diskarray) and the
      > > > > various controller cards...
      > > >
      > > > I have 2 x 200GB Seagate HDD's on ICH5 for quite some time.. Since SATA
      > > > for the ICH% stabilized only recently i had been using lot of patches
      > > > to achieve the functionality. I never had any problem, except that i
      > > > had a failed 160GB Western Digital SATA drive .. Has been working for
      > > > at least 8 months i think, for me, no problems other than the failure..
      > > >
      > > >
      > > > The important point with SATA is that to have SATA support you have to
      > > > have libata support. It depends upon which kernel you are using, the
      > > > complexity of the procedure .. since when you have different patches,
      > > > some patches would have lot of rejects and you would have to patch it
      > > > manually by hand ..
      > > >
      > > >
      > > > I think Brad is having 6 SATA drives, maybe he could explain his
      > > > experience ..
      > > >
      > > > Manu
      > > >
      > > > > cGull.
      > > > > (previously known as mattywix@...)
      > > > >
      > > > > Rajeev wrote on 31.12.2004, 12:47:57:
      > > > > > Dear Friends,
      > > > > >
      > > > > > Happy New Year
      > > > > >
      > > > > > I need to build a Linux server with 1 terra byte (200gb X 6 drives
      >
      > in
      >
      > > > > > raid 5) capacity. I would like to do this with SATA drives. Has any
      > > > > > one done this before? (SATA drives and raid 5)?
      > > >
      > > > No problems at all.. with recent kernels, most of the controllers are
      > > > supported ..
      > > >
      > > > > > Sharing your valuable experience will be appreciated.
      > > > > >
      > > > > > Thanks
      > > > > >
      > > > > > Rajeev
      > > > > >
      > > > > >
      > > > > >
      > > > > >
      > > > > >
      > > > > >
      > > > > >
      > > > > > Yahoo! Groups Links
      > > > > >
      > > > > > To visit your group on the web, go to:
      > > > > > http://groups.yahoo.com/group/dubailug/
      > > > > >
      > > > > > To unsubscribe from this group, send an email to:
      > > > > > dubailug-unsubscribe@yahoogroups.com
      > > > > >
      > > > > > Your use of Yahoo! Groups is subject to:
      > > > > > http://docs.yahoo.com/info/terms/
      > > > >
      > > > > Yahoo! Groups Links
      > > >
      > > > Yahoo! Groups Links
      > > >
      > > > To visit your group on the web, go to:
      > > > http://groups.yahoo.com/group/dubailug/
      > > >
      > > > To unsubscribe from this group, send an email to:
      > > > dubailug-unsubscribe@yahoogroups.com
      > > >
      > > > Your use of Yahoo! Groups is subject to:
      > > > http://docs.yahoo.com/info/terms/
      > >
      > > Yahoo! Groups Links
      >
      > Yahoo! Groups Links
      >
      >
      >
      >
      >
      >
      >
      >
      >
      >
      > Yahoo! Groups Links
      >
      >
      >
    • cgull@cgullcharlie.com
      As Manu said, there are many permutations. Here are a few more thoughts... First, If you can live without the manageability of rack hardware, using standard PC
      Message 2 of 29 , Jan 1, 2005
        As Manu said, there are many permutations.

        Here are a few more thoughts...

        First, If you can live without the manageability of rack hardware, using
        standard PC towers saves a lot of cash.

        Second, You say you need 1TB on RAID5 but I didnt see a mention of why
        (or maybe I misread it).
        If you need a contigous 1TB disk space, you can achieve that in other
        ways. For example you can group a bunch of standard cheap IDE drives
        into one large single Logical Partition.

        Alternatively, to get RAID5 cheap, you can buy a PCI card that lets you
        create raid from standard IDE´s without requiring SCSI disks.
        I cant remember the name, but when Im back in Dubai in a few days I can
        let you know who produces these cards.

        Lastly, RAID5 should be faster because instead of writing a file in one
        long linear access to a single disk, you are writing parts of it to
        different disks, so its a slightly more parrallel operation.
        However if its Linux software RAID rather than hardware then I cant tell
        you what the result will be as I have never tried that.

        By the way, as Manu said, you need to clarify why you are using Squid.
        I had it running on a firewall to log user activity, but since realised
        that I was more interested in bandwidth quotas and uninstalled it.
        Squid did have a negative effect on users perception of our network
        performance, which is funny because the idea of caching should be to
        boost perceived performance. I think it was the proxy side of it that
        slowed things down.

        cGull.
        (previously known as mattywix@...)

        Manu Abraham <manu@...> wrote on 01.01.2005, 12:13:57:
        >
        > On Sat January 1 2005 9:25 am, Rajeev wrote:
        > > Manu/ cGull
        > >
        > > I will be purchasing the hardware this week. Could you please suggest me a
        > > guideline spec if I use this machine for the following.
        > >
        > > Fedora C-2 or RH 9
        > Go for the latest FC3. The newer 2.6.x kernel really do give a general
        > performance boost..
        >
        > >
        > > It works mainly as storage server. (SAMBA) (Backup to the main server -
        > > i.e., every night the main server will copy the data to this server and no
        > > user will be directly accessing this one - and it should be maximum cost
        > > effective)
        > >
        > > It works also as Squid Proxy server for below 100 users.
        > >
        > Fast HDD are advised if you intend to do SQUID Object Caching .. In that case
        > i would advise U320 SCA HDD's depending upon the users usage pattern
        >
        > For a proxy, having the Object cache on a RAID array/disk is meaningless.
        > There is too much overhead. If you are looking at putting the Object cache on
        > a RAID array/disk (if you have a dedicated controller, you can consider it as
        > a RAID disk, rather than an array.. You have only /dev/sdxx, instead..), i
        > would suggest using RAID 1(Mirrorring and Duplexing) as this will have a
        > larger performance, since the writes are alternate. RAID 5 gives you only
        > data transfer rate same as single disk
        > whereas RAID 1 will give you very high availability.(If you are looking at
        > RAID 1, i would strongly recomend a Hardware implementation.)
        >
        > Moreover, with a RAID 5 array, rebuilding would be more tedious in the case of
        > a disk failure
        >
        > In this case i would suggest moving onto Hardware RAID solutions rather than a
        > software RAID solution. The Software RAID solution makes use of the CPU's
        > free processor cycles for replication.
        >
        > Whereas in a Hardware RAID solution, there is a dedicated processor and
        > associated memory on the RAID board to achieve that.
        >
        > Maybe if you are cash strapped, you can look out for Zero channel RAID
        > solutions as well which are still cost effective but still better than
        > Software RAID.
        >
        >
        > If that is the case i can suggest you some ASUS SMP server board with a zero
        > channel/Full blown RAID, if you plan to build your own server. These boards
        > are not very expensive and are quite reliable.. 5 years back, i used have
        > have 2 machines, with dual PIII Xeon 800 Mhz with 512MB ECC SDRAM, with an
        > Adaptec I20 Hardware RAID controller with U160 disk's, running RH6.0 and
        > patches having about 120 users, and 146 PIII 800 MHz Machines as PVM cluster
        > nodes and 17 SGI Visualization Workstations, for automobile modelling
        > (CAE/CAM) purposes. That system is still running _unattended_, the reason
        > there is a huge 40KVA UPS system with a battery room, that can last for 2
        > days. I still give the guys a few pointers over phone ... The only trouble
        > the guys compain is when the leased lines go down, when somebody breaks the
        > roads..
        >
        > This machine was used as a mail server, SMB/NFS File server, Doubled as a
        > DHCP/DNS, and a PVM cluster server ..
        >
        >
        > > This will be a machine as you described - no monitor, no keyboard and no
        > > mouse. I will have 1 GB of ram. And I need 1 terra byte space in Raid 5.
        > > Since this machine will be part of a production network I can't do too many
        >
        > Put OS onto separate HDD's itself, if you are planning for extreme low
        > downtime
        >
        > And user data into the RAID disk's/array.
        >
        > That kind of combinations depends upon what level of HA you are looking at
        >
        > For example, for my HA applications, i have a dedicated Hardware RAID U320
        > controller with 2 raid arrays A & B..
        >
        > In my case it is not the number of users, but the bandwidth that is
        > generated .. I have around 25 users only on that HA system as opposed to 100
        > in your case.. And lot of financial data also....... That makes it HA..
        >
        > Array A contains OS and
        > Array B contains User data..
        >
        >
        > Another major suggestion that i can give you on your N/W design is that most
        > probably you will not require to have a proxy...
        >
        > You can get around using Netfilter and NAT.. That will bring down a lot of
        > your requirements... Then that machine would be a firewall machine doing NAT
        > and PAT, but again it is not advisable to use a firewalling machine as your
        > intranet server.
        >
        > Well, there are lot of points to be considered, like what exactly you are
        > trying to achieve, the permutations and combinations are huge ..
        >
        > Manu
        > > experiments.
        > >
        > > I would really appreciate your help,
        > >
        > >
        > >
        > >
        > > -----Original Message-----
        > > From: Manu Abraham [mailto:manu@...]
        > > Sent: 31 December 2004 22:53
        > > To: dubailug@yahoogroups.com
        > > Subject: Re: [dubailug] Sata Drives
        > >
        > > On Fri December 31 2004 10:14 pm, cgull@... wrote:
        > > > Manu,
        > > >
        > > > Thanks for the cool info!
        > > > Can you tell us if it is possible to load this support via a module (for
        > >
        > > It depends upon the kernel .. I had SATA support into 2.4 also One of my
        > > 2.4 machines which i use as my NFS/SAMBA fileserver, has around a total of
        > > > 200Gigs n/w storage on SATA, i have another IBM one with U320 SCA SCSI
        > > also.
        > >
        > > No problems also .. If somebody requires i can compile the kernel modules
        > > if i
        > > have that kernel version. But rather than picking up my binary modules i
        > > would really suggest guys to put in the additional pain of recompiling.. It
        > > is really worth it ..
        > >
        > > > example via insmod) rather than recompile kernel with changed options?
        > >
        > > Once compiled libata can be loaded as a module as i have loaded..
        > > You need to have to enable SCSI support also to have SATA functioning ..
        > >
        > >
        > > appletalk 33972 2
        > > ax25 58476 2
        > > ipx 26540 2
        > > ipv6 237440 18
        > > intel_mch_agp 10128 1
        > > agpgart 30888 2 intel_mch_agp
        > > bluetooth 45924 6 l2cap,hci_vhci,hci_uart,hci_usb
        > > af_packet 20872 2
        > > 3c59x 36904 0
        > > binfmt_misc 10120 1
        > > ata_piix 7684 0
        > > libata 38016 1 ata_piix,[permanent]
        > > sd_mod 20224 0
        > > scsi_mod 108876 5 sg,st,sr_mod,libata,sd_mod
        > >
        > >
        > > For older kernels, for libata support the __only__ way is to patch the
        > > kernel
        > > with the libata patch. If somebody wants to find the patch, it is
        > > maintained by Jeff Garzick. You can find the patches on kernel.org. It
        > > depends how old or what kernel it is to know what other patches if any
        > > libata requires.. I had 2.4.24 or 25... I think i had a lot of patches onto
        > > that..
        > >
        > > I had the ck-xx patches, libata patch, lot of patches, i don't recollect..
        > > what all.. That machine is in a corner humming away without any problems at
        > > all as i said for more than 8 months without any interruptions.. Even
        > > during power up and downs. There is no keyboard/monitor/mouse on that..
        > > Anything goes wrong (nothing has gone wrong except power supply blew up
        > > once) somebody
        > > just presses the reset button (in emergency) else restarts through SSH.. As
        > > simple as that..
        > >
        > > I have a quite heavy data transfer on that machine GB's daily as machine
        > > does
        > > storage for some PVR jobs also.
        > >
        > > Regarding new kernels... There is no problem at all, It is like fish
        > > jumping to water.
        > >
        > > The most problematic one was the ICH5, ie, SATA on the Intel boards, but
        > > even
        > > that, the problems are no longer there with the new kernels..
        > >
        > > Kernel-2.6.9 onwards beautiful support is there, but there is exception of
        > > Seagate HDD's on SiL controllers, but only a firmware problem on the drives
        > > as i said earlier ..
        > >
        > > Manu
        > >
        > > > cGull.
        > > >
        > > > Manu Abraham wrote on 31.12.2004, 18:49:14:
        > > > > On Fri December 31 2004 7:30 pm, cgull@... wrote:
        > > > > > Rajeev,
        > > > > >
        > > > > > I have tried SATA drives on HP hardware and I had some problems.
        > > > >
        > > > > Which controller was this.. It is known that Seagate HDD's and SiL
        > > > > Controllers don't get along well, but therre have been reports that
        > >
        > > HDD's
        > >
        > > > > with newer firmware does not have any problems ..
        > > > >
        > > > > > For example Knoppix and Tomsrtbt both hung on boot because of the
        > >
        > > SATA.
        > >
        > > > > libata i believe .. Try using the SATA drives in compatibity mode,
        > > > > eventhough you loose performance, but after installation of the OS,
        > > > > just rebuild the kernel with the relevant patches..
        > > > >
        > > > > > My Mandrake distro has no problems with it however.
        > > > >
        > > > > Mandrake has libata support . :-)
        > > > >
        > > > > > I searched the internet and found there were some issues with SATA.
        > > > > > To avoid this I just went for SCSI which is guaranteed to work with
        > > > > > Linux.
        > > > >
        > > > > SATA also uses the same SCSI commands under linux... In fact for a SATA
        > > > > HDD you have the regular /dev/sdxx for SATA as compared to SCSI>
        > > > >
        > > > > > I know SATA is cheaper, but then I wasnt the one forking out the cash
        > > > > > so I could make the choice of SCSI instead.
        > > > > > You may also have other issues such as SATA not being hot-pluggable
        > > > > > from
        > > > >
        > > > > For SCSI also, unless you have SCA SCSI you cannot Hot swap... SCA SCSI
        > > > > is still more expensive than standard SCSI, and moreover for SCA SCSI
        > >
        > > you
        > >
        > > > > have to have the drive cage to properly Hotswap.. The only difference
        > > > > between SCA SCSI and SCSI is that the power to the SCSI disk comes
        > > > > through the SCA SCSI connector rather than the power connector.. SCA
        > >
        > > SCSI
        > >
        > > > > has no separate power connectors, but they have a controller situated
        > > > > in the drive cage to shutdown the power to the relevant drive ...
        > > > >
        > > > > > your server or disk array.
        > > > > > Tell us what hardware you plan to use (server and diskarray) and the
        > > > > > various controller cards...
        > > > >
        > > > > I have 2 x 200GB Seagate HDD's on ICH5 for quite some time.. Since SATA
        > > > > for the ICH% stabilized only recently i had been using lot of patches
        > > > > to achieve the functionality. I never had any problem, except that i
        > > > > had a failed 160GB Western Digital SATA drive .. Has been working for
        > > > > at least 8 months i think, for me, no problems other than the failure..
        > > > >
        > > > >
        > > > > The important point with SATA is that to have SATA support you have to
        > > > > have libata support. It depends upon which kernel you are using, the
        > > > > complexity of the procedure .. since when you have different patches,
        > > > > some patches would have lot of rejects and you would have to patch it
        > > > > manually by hand ..
        > > > >
        > > > >
        > > > > I think Brad is having 6 SATA drives, maybe he could explain his
        > > > > experience ..
        > > > >
        > > > > Manu
        > > > >
        > > > > > cGull.
        > > > > > (previously known as mattywix@...)
        > > > > >
        > > > > > Rajeev wrote on 31.12.2004, 12:47:57:
        > > > > > > Dear Friends,
        > > > > > >
        > > > > > > Happy New Year
        > > > > > >
        > > > > > > I need to build a Linux server with 1 terra byte (200gb X 6 drives
        > >
        > > in
        > >
        > > > > > > raid 5) capacity. I would like to do this with SATA drives. Has any
        > > > > > > one done this before? (SATA drives and raid 5)?
        > > > >
        > > > > No problems at all.. with recent kernels, most of the controllers are
        > > > > supported ..
        > > > >
        > > > > > > Sharing your valuable experience will be appreciated.
        > > > > > >
        > > > > > > Thanks
        > > > > > >
        > > > > > > Rajeev
        > > > > > >
        > > > > > >
        > > > > > >
        > > > > > >
        > > > > > >
        > > > > > >
        > > > > > >
        > > > > > > Yahoo! Groups Links
        > > > > > >
        > > > > > > To visit your group on the web, go to:
        > > > > > > http://groups.yahoo.com/group/dubailug/
        > > > > > >
        > > > > > > To unsubscribe from this group, send an email to:
        > > > > > > dubailug-unsubscribe@yahoogroups.com
        > > > > > >
        > > > > > > Your use of Yahoo! Groups is subject to:
        > > > > > > http://docs.yahoo.com/info/terms/
        > > > > >
        > > > > > Yahoo! Groups Links
        > > > >
        > > > > Yahoo! Groups Links
        > > > >
        > > > > To visit your group on the web, go to:
        > > > > http://groups.yahoo.com/group/dubailug/
        > > > >
        > > > > To unsubscribe from this group, send an email to:
        > > > > dubailug-unsubscribe@yahoogroups.com
        > > > >
        > > > > Your use of Yahoo! Groups is subject to:
        > > > > http://docs.yahoo.com/info/terms/
        > > >
        > > > Yahoo! Groups Links
        > >
        > > Yahoo! Groups Links
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > >
        > > Yahoo! Groups Links
        > >
        > >
        > >
        >
        >
        >
        > Yahoo! Groups Links
        >
        > To visit your group on the web, go to:
        > http://groups.yahoo.com/group/dubailug/
        >
        > To unsubscribe from this group, send an email to:
        > dubailug-unsubscribe@yahoogroups.com
        >
        > Your use of Yahoo! Groups is subject to:
        > http://docs.yahoo.com/info/terms/
      • Manu Abraham
        ... If 1TB required as contiguous section, better not use EXT3, but rather use XFS. ... Large filesystems on software RAID5 is no joke.. The pain starts when
        Message 3 of 29 , Jan 1, 2005
          On Sat January 1 2005 6:26 pm, cgull@... wrote:
          > As Manu said, there are many permutations.
          >
          > Here are a few more thoughts...
          >
          > First, If you can live without the manageability of rack hardware, using
          > standard PC towers saves a lot of cash.
          >
          > Second, You say you need 1TB on RAID5 but I didnt see a mention of why
          > (or maybe I misread it).
          > If you need a contigous 1TB disk space, you can achieve that in other
          > ways. For example you can group a bunch of standard cheap IDE drives
          > into one large single Logical Partition.
          If 1TB required as contiguous section, better not use EXT3, but rather use
          XFS.

          >
          > Alternatively, to get RAID5 cheap, you can buy a PCI card that lets you
          > create raid from standard IDE´s without requiring SCSI disks.
          > I cant remember the name, but when Im back in Dubai in a few days I can
          > let you know who produces these cards.
          >
          > Lastly, RAID5 should be faster because instead of writing a file in one
          > long linear access to a single disk, you are writing parts of it to
          > different disks, so its a slightly more parrallel operation.
          > However if its Linux software RAID rather than hardware then I cant tell
          > you what the result will be as I have never tried that.
          Large filesystems on software RAID5 is no joke.. The pain starts when
          rebuilding when one of the disks in the stripes fail.. Maybe sometimes build
          fails also... I think people who had RAID rebuild's after a failure can chip
          in ..

          That is the point where hardware RAID comes into the picture.. You are up back
          again in no time .. Wow .. saved me once .. never had a second experience ..
          >
          > By the way, as Manu said, you need to clarify why you are using Squid.
          > I had it running on a firewall to log user activity, but since realised
          > that I was more interested in bandwidth quotas and uninstalled it.
          > Squid did have a negative effect on users perception of our network
          > performance, which is funny because the idea of caching should be to
          > boost perceived performance. I think it was the proxy side of it that
          > slowed things down.

          When you have the object cache on a slower HDD, like an IDE and so on (one
          should not forget that the cached data also is not small, you can change the
          size though), it would be faster if the packet was retrieved from source
          rather than from the object cache, in certain cases when the HDD really
          starts grinding. That's why good caching systems have very fast HDD's with
          large buffers on the HDD's and high speed SCSI HDD's in the past..

          Therefore it is advised that caching systems have large physical memory for
          the buffering..

          Manu
          >
          > cGull.
          > (previously known as mattywix@...)
          >
          > Manu Abraham <manu@...> wrote on 01.01.2005, 12:13:57:
          > > On Sat January 1 2005 9:25 am, Rajeev wrote:
          > > > Manu/ cGull
          > > >
          > > > I will be purchasing the hardware this week. Could you please suggest
          > > > me a guideline spec if I use this machine for the following.
          > > >
          > > > Fedora C-2 or RH 9
          > >
          > > Go for the latest FC3. The newer 2.6.x kernel really do give a general
          > > performance boost..
          > >
          > > > It works mainly as storage server. (SAMBA) (Backup to the main server -
          > > > i.e., every night the main server will copy the data to this server and
          > > > no user will be directly accessing this one - and it should be maximum
          > > > cost effective)
          > > >
          > > > It works also as Squid Proxy server for below 100 users.
          > >
          > > Fast HDD are advised if you intend to do SQUID Object Caching .. In that
          > > case i would advise U320 SCA HDD's depending upon the users usage pattern
          > >
          > > For a proxy, having the Object cache on a RAID array/disk is meaningless.
          > > There is too much overhead. If you are looking at putting the Object
          > > cache on a RAID array/disk (if you have a dedicated controller, you can
          > > consider it as a RAID disk, rather than an array.. You have only
          > > /dev/sdxx, instead..), i would suggest using RAID 1(Mirrorring and
          > > Duplexing) as this will have a larger performance, since the writes are
          > > alternate. RAID 5 gives you only data transfer rate same as single disk
          > > whereas RAID 1 will give you very high availability.(If you are looking
          > > at RAID 1, i would strongly recomend a Hardware implementation.)
          > >
          > > Moreover, with a RAID 5 array, rebuilding would be more tedious in the
          > > case of a disk failure
          > >
          > > In this case i would suggest moving onto Hardware RAID solutions rather
          > > than a software RAID solution. The Software RAID solution makes use of
          > > the CPU's free processor cycles for replication.
          > >
          > > Whereas in a Hardware RAID solution, there is a dedicated processor and
          > > associated memory on the RAID board to achieve that.
          > >
          > > Maybe if you are cash strapped, you can look out for Zero channel RAID
          > > solutions as well which are still cost effective but still better than
          > > Software RAID.
          > >
          > >
          > > If that is the case i can suggest you some ASUS SMP server board with a
          > > zero channel/Full blown RAID, if you plan to build your own server. These
          > > boards are not very expensive and are quite reliable.. 5 years back, i
          > > used have have 2 machines, with dual PIII Xeon 800 Mhz with 512MB ECC
          > > SDRAM, with an Adaptec I20 Hardware RAID controller with U160 disk's,
          > > running RH6.0 and patches having about 120 users, and 146 PIII 800 MHz
          > > Machines as PVM cluster nodes and 17 SGI Visualization Workstations, for
          > > automobile modelling (CAE/CAM) purposes. That system is still running
          > > _unattended_, the reason there is a huge 40KVA UPS system with a battery
          > > room, that can last for 2 days. I still give the guys a few pointers
          > > over phone ... The only trouble the guys compain is when the leased lines
          > > go down, when somebody breaks the roads..
          > >
          > > This machine was used as a mail server, SMB/NFS File server, Doubled as a
          > > DHCP/DNS, and a PVM cluster server ..
          > >
          > > > This will be a machine as you described - no monitor, no keyboard and
          > > > no mouse. I will have 1 GB of ram. And I need 1 terra byte space in
          > > > Raid 5. Since this machine will be part of a production network I can't
          > > > do too many
          > >
          > > Put OS onto separate HDD's itself, if you are planning for extreme low
          > > downtime
          > >
          > > And user data into the RAID disk's/array.
          > >
          > > That kind of combinations depends upon what level of HA you are looking
          > > at
          > >
          > > For example, for my HA applications, i have a dedicated Hardware RAID
          > > U320 controller with 2 raid arrays A & B..
          > >
          > > In my case it is not the number of users, but the bandwidth that is
          > > generated .. I have around 25 users only on that HA system as opposed to
          > > 100 in your case.. And lot of financial data also....... That makes it
          > > HA..
          > >
          > > Array A contains OS and
          > > Array B contains User data..
          > >
          > >
          > > Another major suggestion that i can give you on your N/W design is that
          > > most probably you will not require to have a proxy...
          > >
          > > You can get around using Netfilter and NAT.. That will bring down a lot
          > > of your requirements... Then that machine would be a firewall machine
          > > doing NAT and PAT, but again it is not advisable to use a firewalling
          > > machine as your intranet server.
          > >
          > > Well, there are lot of points to be considered, like what exactly you are
          > > trying to achieve, the permutations and combinations are huge ..
          > >
          > > Manu
          > >
          > > > experiments.
          > > >
          > > > I would really appreciate your help,
          > > >
          > > >
          > > >
          > > >
          > > > -----Original Message-----
          > > > From: Manu Abraham [mailto:manu@...]
          > > > Sent: 31 December 2004 22:53
          > > > To: dubailug@yahoogroups.com
          > > > Subject: Re: [dubailug] Sata Drives
          > > >
          > > > On Fri December 31 2004 10:14 pm, cgull@... wrote:
          > > > > Manu,
          > > > >
          > > > > Thanks for the cool info!
          > > > > Can you tell us if it is possible to load this support via a module
          > > > > (for
          > > >
          > > > It depends upon the kernel .. I had SATA support into 2.4 also One of
          > > > my 2.4 machines which i use as my NFS/SAMBA fileserver, has around a
          > > > total of
          > > >
          > > > > 200Gigs n/w storage on SATA, i have another IBM one with U320 SCA
          > > > > SCSI
          > > >
          > > > also.
          > > >
          > > > No problems also .. If somebody requires i can compile the kernel
          > > > modules if i
          > > > have that kernel version. But rather than picking up my binary modules
          > > > i would really suggest guys to put in the additional pain of
          > > > recompiling.. It is really worth it ..
          > > >
          > > > > example via insmod) rather than recompile kernel with changed
          > > > > options?
          > > >
          > > > Once compiled libata can be loaded as a module as i have loaded..
          > > > You need to have to enable SCSI support also to have SATA functioning
          > > > ..
          > > >
          > > >
          > > > appletalk 33972 2
          > > > ax25 58476 2
          > > > ipx 26540 2
          > > > ipv6 237440 18
          > > > intel_mch_agp 10128 1
          > > > agpgart 30888 2 intel_mch_agp
          > > > bluetooth 45924 6 l2cap,hci_vhci,hci_uart,hci_usb
          > > > af_packet 20872 2
          > > > 3c59x 36904 0
          > > > binfmt_misc 10120 1
          > > > ata_piix 7684 0
          > > > libata 38016 1 ata_piix,[permanent]
          > > > sd_mod 20224 0
          > > > scsi_mod 108876 5 sg,st,sr_mod,libata,sd_mod
          > > >
          > > >
          > > > For older kernels, for libata support the __only__ way is to patch the
          > > > kernel
          > > > with the libata patch. If somebody wants to find the patch, it is
          > > > maintained by Jeff Garzick. You can find the patches on kernel.org. It
          > > > depends how old or what kernel it is to know what other patches if any
          > > > libata requires.. I had 2.4.24 or 25... I think i had a lot of patches
          > > > onto that..
          > > >
          > > > I had the ck-xx patches, libata patch, lot of patches, i don't
          > > > recollect.. what all.. That machine is in a corner humming away without
          > > > any problems at all as i said for more than 8 months without any
          > > > interruptions.. Even during power up and downs. There is no
          > > > keyboard/monitor/mouse on that.. Anything goes wrong (nothing has gone
          > > > wrong except power supply blew up once) somebody
          > > > just presses the reset button (in emergency) else restarts through
          > > > SSH.. As simple as that..
          > > >
          > > > I have a quite heavy data transfer on that machine GB's daily as
          > > > machine does
          > > > storage for some PVR jobs also.
          > > >
          > > > Regarding new kernels... There is no problem at all, It is like fish
          > > > jumping to water.
          > > >
          > > > The most problematic one was the ICH5, ie, SATA on the Intel boards,
          > > > but even
          > > > that, the problems are no longer there with the new kernels..
          > > >
          > > > Kernel-2.6.9 onwards beautiful support is there, but there is exception
          > > > of Seagate HDD's on SiL controllers, but only a firmware problem on the
          > > > drives as i said earlier ..
          > > >
          > > > Manu
          > > >
          > > > > cGull.
          > > > >
          > > > > Manu Abraham wrote on 31.12.2004, 18:49:14:
          > > > > > On Fri December 31 2004 7:30 pm, cgull@... wrote:
          > > > > > > Rajeev,
          > > > > > >
          > > > > > > I have tried SATA drives on HP hardware and I had some problems.
          > > > > >
          > > > > > Which controller was this.. It is known that Seagate HDD's and SiL
          > > > > > Controllers don't get along well, but therre have been reports that
          > > >
          > > > HDD's
          > > >
          > > > > > with newer firmware does not have any problems ..
          > > > > >
          > > > > > > For example Knoppix and Tomsrtbt both hung on boot because of the
          > > >
          > > > SATA.
          > > >
          > > > > > libata i believe .. Try using the SATA drives in compatibity mode,
          > > > > > eventhough you loose performance, but after installation of the OS,
          > > > > > just rebuild the kernel with the relevant patches..
          > > > > >
          > > > > > > My Mandrake distro has no problems with it however.
          > > > > >
          > > > > > Mandrake has libata support . :-)
          > > > > >
          > > > > > > I searched the internet and found there were some issues with
          > > > > > > SATA. To avoid this I just went for SCSI which is guaranteed to
          > > > > > > work with Linux.
          > > > > >
          > > > > > SATA also uses the same SCSI commands under linux... In fact for a
          > > > > > SATA HDD you have the regular /dev/sdxx for SATA as compared to
          > > > > > SCSI>
          > > > > >
          > > > > > > I know SATA is cheaper, but then I wasnt the one forking out the
          > > > > > > cash so I could make the choice of SCSI instead.
          > > > > > > You may also have other issues such as SATA not being
          > > > > > > hot-pluggable from
          > > > > >
          > > > > > For SCSI also, unless you have SCA SCSI you cannot Hot swap... SCA
          > > > > > SCSI is still more expensive than standard SCSI, and moreover for
          > > > > > SCA SCSI
          > > >
          > > > you
          > > >
          > > > > > have to have the drive cage to properly Hotswap.. The only
          > > > > > difference between SCA SCSI and SCSI is that the power to the SCSI
          > > > > > disk comes through the SCA SCSI connector rather than the power
          > > > > > connector.. SCA
          > > >
          > > > SCSI
          > > >
          > > > > > has no separate power connectors, but they have a controller
          > > > > > situated in the drive cage to shutdown the power to the relevant
          > > > > > drive ...
          > > > > >
          > > > > > > your server or disk array.
          > > > > > > Tell us what hardware you plan to use (server and diskarray) and
          > > > > > > the various controller cards...
          > > > > >
          > > > > > I have 2 x 200GB Seagate HDD's on ICH5 for quite some time.. Since
          > > > > > SATA for the ICH% stabilized only recently i had been using lot of
          > > > > > patches to achieve the functionality. I never had any problem,
          > > > > > except that i had a failed 160GB Western Digital SATA drive .. Has
          > > > > > been working for at least 8 months i think, for me, no problems
          > > > > > other than the failure..
          > > > > >
          > > > > >
          > > > > > The important point with SATA is that to have SATA support you have
          > > > > > to have libata support. It depends upon which kernel you are using,
          > > > > > the complexity of the procedure .. since when you have different
          > > > > > patches, some patches would have lot of rejects and you would have
          > > > > > to patch it manually by hand ..
          > > > > >
          > > > > >
          > > > > > I think Brad is having 6 SATA drives, maybe he could explain his
          > > > > > experience ..
          > > > > >
          > > > > > Manu
          > > > > >
          > > > > > > cGull.
          > > > > > > (previously known as mattywix@...)
          > > > > > >
          > > > > > > Rajeev wrote on 31.12.2004, 12:47:57:
          > > > > > > > Dear Friends,
          > > > > > > >
          > > > > > > > Happy New Year
          > > > > > > >
          > > > > > > > I need to build a Linux server with 1 terra byte (200gb X 6
          > > > > > > > drives
          > > >
          > > > in
          > > >
          > > > > > > > raid 5) capacity. I would like to do this with SATA drives. Has
          > > > > > > > any one done this before? (SATA drives and raid 5)?
          > > > > >
          > > > > > No problems at all.. with recent kernels, most of the controllers
          > > > > > are supported ..
          > > > > >
          > > > > > > > Sharing your valuable experience will be appreciated.
          > > > > > > >
          > > > > > > > Thanks
          > > > > > > >
          > > > > > > > Rajeev
          > > > > > > >
          > > > > > > >
          > > > > > > >
          > > > > > > >
          > > > > > > >
          > > > > > > >
          > > > > > > >
          > > > > > > > Yahoo! Groups Links
          > > > > > > >
          > > > > > > > To visit your group on the web, go to:
          > > > > > > > http://groups.yahoo.com/group/dubailug/
          > > > > > > >
          > > > > > > > To unsubscribe from this group, send an email to:
          > > > > > > > dubailug-unsubscribe@yahoogroups.com
          > > > > > > >
          > > > > > > > Your use of Yahoo! Groups is subject to:
          > > > > > > > http://docs.yahoo.com/info/terms/
          > > > > > >
          > > > > > > Yahoo! Groups Links
          > > > > >
          > > > > > Yahoo! Groups Links
          > > > > >
          > > > > > To visit your group on the web, go to:
          > > > > > http://groups.yahoo.com/group/dubailug/
          > > > > >
          > > > > > To unsubscribe from this group, send an email to:
          > > > > > dubailug-unsubscribe@yahoogroups.com
          > > > > >
          > > > > > Your use of Yahoo! Groups is subject to:
          > > > > > http://docs.yahoo.com/info/terms/
          > > > >
          > > > > Yahoo! Groups Links
          > > >
          > > > Yahoo! Groups Links
          > > >
          > > >
          > > >
          > > >
          > > >
          > > >
          > > >
          > > >
          > > >
          > > >
          > > > Yahoo! Groups Links
          > >
          > > Yahoo! Groups Links
          > >
          > > To visit your group on the web, go to:
          > > http://groups.yahoo.com/group/dubailug/
          > >
          > > To unsubscribe from this group, send an email to:
          > > dubailug-unsubscribe@yahoogroups.com
          > >
          > > Your use of Yahoo! Groups is subject to:
          > > http://docs.yahoo.com/info/terms/
          >
          > Yahoo! Groups Links
          >
          >
          >
        • Rajeev
          Dear Manu /cGull Thanks for your support. I need 1 TB because this machine is used for backup storage purpose. I will be storing large PSD files
          Message 4 of 29 , Jan 1, 2005
            Dear Manu /cGull

            Thanks for your support.

            I need 1 TB because this machine is used for backup storage purpose. I will
            be storing large PSD files

            http://www.adaptec.com/worldwide/product/proddetail.html?sess=no&language=En
            glish+US&prodkey=AAR-2810SA&cat=%2fTechnology%2fSerial+ATA%2fRAID+Controller
            s


            I think the above raid card will solve the purpose. What do you think?




            Regards

            Rajeev







            -----Original Message-----
            From: Manu Abraham [mailto:manu@...]
            Sent: 01 January 2005 15:14
            To: dubailug@yahoogroups.com
            Subject: Re: [dubailug] Sata Drives


            On Sat January 1 2005 9:25 am, Rajeev wrote:
            > Manu/ cGull
            >
            > I will be purchasing the hardware this week. Could you please suggest me a
            > guideline spec if I use this machine for the following.
            >
            > Fedora C-2 or RH 9
            Go for the latest FC3. The newer 2.6.x kernel really do give a general
            performance boost..

            >
            > It works mainly as storage server. (SAMBA) (Backup to the main server -
            > i.e., every night the main server will copy the data to this server and no
            > user will be directly accessing this one - and it should be maximum cost
            > effective)
            >
            > It works also as Squid Proxy server for below 100 users.
            >
            Fast HDD are advised if you intend to do SQUID Object Caching .. In that
            case
            i would advise U320 SCA HDD's depending upon the users usage pattern

            For a proxy, having the Object cache on a RAID array/disk is meaningless.
            There is too much overhead. If you are looking at putting the Object cache
            on
            a RAID array/disk (if you have a dedicated controller, you can consider it
            as
            a RAID disk, rather than an array.. You have only /dev/sdxx, instead..), i
            would suggest using RAID 1(Mirrorring and Duplexing) as this will have a
            larger performance, since the writes are alternate. RAID 5 gives you only
            data transfer rate same as single disk
            whereas RAID 1 will give you very high availability.(If you are looking at
            RAID 1, i would strongly recomend a Hardware implementation.)

            Moreover, with a RAID 5 array, rebuilding would be more tedious in the case
            of
            a disk failure

            In this case i would suggest moving onto Hardware RAID solutions rather than
            a
            software RAID solution. The Software RAID solution makes use of the CPU's
            free processor cycles for replication.

            Whereas in a Hardware RAID solution, there is a dedicated processor and
            associated memory on the RAID board to achieve that.

            Maybe if you are cash strapped, you can look out for Zero channel RAID
            solutions as well which are still cost effective but still better than
            Software RAID.


            If that is the case i can suggest you some ASUS SMP server board with a zero
            channel/Full blown RAID, if you plan to build your own server. These boards
            are not very expensive and are quite reliable.. 5 years back, i used have
            have 2 machines, with dual PIII Xeon 800 Mhz with 512MB ECC SDRAM, with an
            Adaptec I20 Hardware RAID controller with U160 disk's, running RH6.0 and
            patches having about 120 users, and 146 PIII 800 MHz Machines as PVM cluster
            nodes and 17 SGI Visualization Workstations, for automobile modelling
            (CAE/CAM) purposes. That system is still running _unattended_, the reason
            there is a huge 40KVA UPS system with a battery room, that can last for 2
            days. I still give the guys a few pointers over phone ... The only trouble
            the guys compain is when the leased lines go down, when somebody breaks the
            roads..

            This machine was used as a mail server, SMB/NFS File server, Doubled as a
            DHCP/DNS, and a PVM cluster server ..


            > This will be a machine as you described - no monitor, no keyboard and no
            > mouse. I will have 1 GB of ram. And I need 1 terra byte space in Raid 5.
            > Since this machine will be part of a production network I can't do too
            many

            Put OS onto separate HDD's itself, if you are planning for extreme low
            downtime

            And user data into the RAID disk's/array.

            That kind of combinations depends upon what level of HA you are looking at

            For example, for my HA applications, i have a dedicated Hardware RAID U320
            controller with 2 raid arrays A & B..

            In my case it is not the number of users, but the bandwidth that is
            generated .. I have around 25 users only on that HA system as opposed to 100
            in your case.. And lot of financial data also....... That makes it HA..

            Array A contains OS and
            Array B contains User data..


            Another major suggestion that i can give you on your N/W design is that most
            probably you will not require to have a proxy...

            You can get around using Netfilter and NAT.. That will bring down a lot of
            your requirements... Then that machine would be a firewall machine doing NAT
            and PAT, but again it is not advisable to use a firewalling machine as your
            intranet server.

            Well, there are lot of points to be considered, like what exactly you are
            trying to achieve, the permutations and combinations are huge ..

            Manu
            > experiments.
            >
            > I would really appreciate your help,
            >
            >
            >
            >
            > -----Original Message-----
            > From: Manu Abraham [mailto:manu@...]
            > Sent: 31 December 2004 22:53
            > To: dubailug@yahoogroups.com
            > Subject: Re: [dubailug] Sata Drives
            >
            > On Fri December 31 2004 10:14 pm, cgull@... wrote:
            > > Manu,
            > >
            > > Thanks for the cool info!
            > > Can you tell us if it is possible to load this support via a module (for
            >
            > It depends upon the kernel .. I had SATA support into 2.4 also One of my
            > 2.4 machines which i use as my NFS/SAMBA fileserver, has around a total of
            > > 200Gigs n/w storage on SATA, i have another IBM one with U320 SCA SCSI
            > also.
            >
            > No problems also .. If somebody requires i can compile the kernel modules
            > if i
            > have that kernel version. But rather than picking up my binary modules i
            > would really suggest guys to put in the additional pain of recompiling..
            It
            > is really worth it ..
            >
            > > example via insmod) rather than recompile kernel with changed options?
            >
            > Once compiled libata can be loaded as a module as i have loaded..
            > You need to have to enable SCSI support also to have SATA functioning ..
            >
            >
            > appletalk 33972 2
            > ax25 58476 2
            > ipx 26540 2
            > ipv6 237440 18
            > intel_mch_agp 10128 1
            > agpgart 30888 2 intel_mch_agp
            > bluetooth 45924 6 l2cap,hci_vhci,hci_uart,hci_usb
            > af_packet 20872 2
            > 3c59x 36904 0
            > binfmt_misc 10120 1
            > ata_piix 7684 0
            > libata 38016 1 ata_piix,[permanent]
            > sd_mod 20224 0
            > scsi_mod 108876 5 sg,st,sr_mod,libata,sd_mod
            >
            >
            > For older kernels, for libata support the __only__ way is to patch the
            > kernel
            > with the libata patch. If somebody wants to find the patch, it is
            > maintained by Jeff Garzick. You can find the patches on kernel.org. It
            > depends how old or what kernel it is to know what other patches if any
            > libata requires.. I had 2.4.24 or 25... I think i had a lot of patches
            onto
            > that..
            >
            > I had the ck-xx patches, libata patch, lot of patches, i don't recollect..
            > what all.. That machine is in a corner humming away without any problems
            at
            > all as i said for more than 8 months without any interruptions.. Even
            > during power up and downs. There is no keyboard/monitor/mouse on that..
            > Anything goes wrong (nothing has gone wrong except power supply blew up
            > once) somebody
            > just presses the reset button (in emergency) else restarts through SSH..
            As
            > simple as that..
            >
            > I have a quite heavy data transfer on that machine GB's daily as machine
            > does
            > storage for some PVR jobs also.
            >
            > Regarding new kernels... There is no problem at all, It is like fish
            > jumping to water.
            >
            > The most problematic one was the ICH5, ie, SATA on the Intel boards, but
            > even
            > that, the problems are no longer there with the new kernels..
            >
            > Kernel-2.6.9 onwards beautiful support is there, but there is exception of
            > Seagate HDD's on SiL controllers, but only a firmware problem on the
            drives
            > as i said earlier ..
            >
            > Manu
            >
            > > cGull.
            > >
            > > Manu Abraham <manu@...> wrote on 31.12.2004, 18:49:14:
            > > > On Fri December 31 2004 7:30 pm, cgull@... wrote:
            > > > > Rajeev,
            > > > >
            > > > > I have tried SATA drives on HP hardware and I had some problems.
            > > >
            > > > Which controller was this.. It is known that Seagate HDD's and SiL
            > > > Controllers don't get along well, but therre have been reports that
            >
            > HDD's
            >
            > > > with newer firmware does not have any problems ..
            > > >
            > > > > For example Knoppix and Tomsrtbt both hung on boot because of the
            >
            > SATA.
            >
            > > > libata i believe .. Try using the SATA drives in compatibity mode,
            > > > eventhough you loose performance, but after installation of the OS,
            > > > just rebuild the kernel with the relevant patches..
            > > >
            > > > > My Mandrake distro has no problems with it however.
            > > >
            > > > Mandrake has libata support . :-)
            > > >
            > > > > I searched the internet and found there were some issues with SATA.
            > > > > To avoid this I just went for SCSI which is guaranteed to work with
            > > > > Linux.
            > > >
            > > > SATA also uses the same SCSI commands under linux... In fact for a
            SATA
            > > > HDD you have the regular /dev/sdxx for SATA as compared to SCSI>
            > > >
            > > > > I know SATA is cheaper, but then I wasnt the one forking out the
            cash
            > > > > so I could make the choice of SCSI instead.
            > > > > You may also have other issues such as SATA not being hot-pluggable
            > > > > from
            > > >
            > > > For SCSI also, unless you have SCA SCSI you cannot Hot swap... SCA
            SCSI
            > > > is still more expensive than standard SCSI, and moreover for SCA SCSI
            >
            > you
            >
            > > > have to have the drive cage to properly Hotswap.. The only difference
            > > > between SCA SCSI and SCSI is that the power to the SCSI disk comes
            > > > through the SCA SCSI connector rather than the power connector.. SCA
            >
            > SCSI
            >
            > > > has no separate power connectors, but they have a controller situated
            > > > in the drive cage to shutdown the power to the relevant drive ...
            > > >
            > > > > your server or disk array.
            > > > > Tell us what hardware you plan to use (server and diskarray) and the
            > > > > various controller cards...
            > > >
            > > > I have 2 x 200GB Seagate HDD's on ICH5 for quite some time.. Since
            SATA
            > > > for the ICH% stabilized only recently i had been using lot of patches
            > > > to achieve the functionality. I never had any problem, except that i
            > > > had a failed 160GB Western Digital SATA drive .. Has been working for
            > > > at least 8 months i think, for me, no problems other than the
            failure..
            > > >
            > > >
            > > > The important point with SATA is that to have SATA support you have to
            > > > have libata support. It depends upon which kernel you are using, the
            > > > complexity of the procedure .. since when you have different patches,
            > > > some patches would have lot of rejects and you would have to patch it
            > > > manually by hand ..
            > > >
            > > >
            > > > I think Brad is having 6 SATA drives, maybe he could explain his
            > > > experience ..
            > > >
            > > > Manu
            > > >
            > > > > cGull.
            > > > > (previously known as mattywix@...)
            > > > >
            > > > > Rajeev wrote on 31.12.2004, 12:47:57:
            > > > > > Dear Friends,
            > > > > >
            > > > > > Happy New Year
            > > > > >
            > > > > > I need to build a Linux server with 1 terra byte (200gb X 6 drives
            >
            > in
            >
            > > > > > raid 5) capacity. I would like to do this with SATA drives. Has
            any
            > > > > > one done this before? (SATA drives and raid 5)?
            > > >
            > > > No problems at all.. with recent kernels, most of the controllers are
            > > > supported ..
            > > >
            > > > > > Sharing your valuable experience will be appreciated.
            > > > > >
            > > > > > Thanks
            > > > > >
            > > > > > Rajeev
            > > > > >
            > > > > >
            > > > > >
            > > > > >
            > > > > >
            > > > > >
            > > > > >
            > > > > > Yahoo! Groups Links
            > > > > >
            > > > > > To visit your group on the web, go to:
            > > > > > http://groups.yahoo.com/group/dubailug/
            > > > > >
            > > > > > To unsubscribe from this group, send an email to:
            > > > > > dubailug-unsubscribe@yahoogroups.com
            > > > > >
            > > > > > Your use of Yahoo! Groups is subject to:
            > > > > > http://docs.yahoo.com/info/terms/
            > > > >
            > > > > Yahoo! Groups Links
            > > >
            > > > Yahoo! Groups Links
            > > >
            > > > To visit your group on the web, go to:
            > > > http://groups.yahoo.com/group/dubailug/
            > > >
            > > > To unsubscribe from this group, send an email to:
            > > > dubailug-unsubscribe@yahoogroups.com
            > > >
            > > > Your use of Yahoo! Groups is subject to:
            > > > http://docs.yahoo.com/info/terms/
            > >
            > > Yahoo! Groups Links
            >
            > Yahoo! Groups Links
            >
            >
            >
            >
            >
            >
            >
            >
            >
            >
            > Yahoo! Groups Links
            >
            >
            >



            Yahoo! Groups Links
          • Manu Abraham
            ... Adaptec cards are generally good in quality and performance, only thing is that it is slightly expensive, but worth it.. BTW that is a 64 bit card, i
            Message 5 of 29 , Jan 1, 2005
              On Sat January 1 2005 9:00 pm, you wrote:
              > Dear Manu /cGull
              >
              > Thanks for your support.
              >
              > I need 1 TB because this machine is used for backup storage purpose. I will
              > be storing large PSD files
              >
              > http://www.adaptec.com/worldwide/product/proddetail.html?sess=no&language=E
              >n
              > glish+US&prodkey=AAR-2810SA&cat=%2fTechnology%2fSerial+ATA%2fRAID+Controlle
              >r s
              >
              >
              > I think the above raid card will solve the purpose. What do you think?
              >
              Adaptec cards are generally good in quality and performance, only thing is
              that it is slightly expensive, but worth it..

              BTW that is a 64 bit card, i think. Could not see the connectors too clearly..
              You require a Xeon processor based machine to use that card in that case..

              Anyway it is better to have 1TB on a 64bit/133MHz PCI bus rather than 32
              bit/33MHz.

              Manu
              >
              >
              >
              > Regards
              >
              > Rajeev
              >
              >
              >
              >
              >
              >
              >
              > -----Original Message-----
              > From: Manu Abraham [mailto:manu@...]
              > Sent: 01 January 2005 15:14
              > To: dubailug@yahoogroups.com
              > Subject: Re: [dubailug] Sata Drives
              >
              > On Sat January 1 2005 9:25 am, Rajeev wrote:
              > > Manu/ cGull
              > >
              > > I will be purchasing the hardware this week. Could you please suggest me
              > > a guideline spec if I use this machine for the following.
              > >
              > > Fedora C-2 or RH 9
              >
              > Go for the latest FC3. The newer 2.6.x kernel really do give a general
              > performance boost..
              >
              > > It works mainly as storage server. (SAMBA) (Backup to the main server -
              > > i.e., every night the main server will copy the data to this server and
              > > no user will be directly accessing this one - and it should be maximum
              > > cost effective)
              > >
              > > It works also as Squid Proxy server for below 100 users.
              >
              > Fast HDD are advised if you intend to do SQUID Object Caching .. In that
              > case
              > i would advise U320 SCA HDD's depending upon the users usage pattern
              >
              > For a proxy, having the Object cache on a RAID array/disk is meaningless.
              > There is too much overhead. If you are looking at putting the Object cache
              > on
              > a RAID array/disk (if you have a dedicated controller, you can consider it
              > as
              > a RAID disk, rather than an array.. You have only /dev/sdxx, instead..), i
              > would suggest using RAID 1(Mirrorring and Duplexing) as this will have a
              > larger performance, since the writes are alternate. RAID 5 gives you only
              > data transfer rate same as single disk
              > whereas RAID 1 will give you very high availability.(If you are looking at
              > RAID 1, i would strongly recomend a Hardware implementation.)
              >
              > Moreover, with a RAID 5 array, rebuilding would be more tedious in the case
              > of
              > a disk failure
              >
              > In this case i would suggest moving onto Hardware RAID solutions rather
              > than a
              > software RAID solution. The Software RAID solution makes use of the CPU's
              > free processor cycles for replication.
              >
              > Whereas in a Hardware RAID solution, there is a dedicated processor and
              > associated memory on the RAID board to achieve that.
              >
              > Maybe if you are cash strapped, you can look out for Zero channel RAID
              > solutions as well which are still cost effective but still better than
              > Software RAID.
              >
              >
              > If that is the case i can suggest you some ASUS SMP server board with a
              > zero channel/Full blown RAID, if you plan to build your own server. These
              > boards are not very expensive and are quite reliable.. 5 years back, i used
              > have have 2 machines, with dual PIII Xeon 800 Mhz with 512MB ECC SDRAM,
              > with an Adaptec I20 Hardware RAID controller with U160 disk's, running
              > RH6.0 and patches having about 120 users, and 146 PIII 800 MHz Machines as
              > PVM cluster nodes and 17 SGI Visualization Workstations, for automobile
              > modelling (CAE/CAM) purposes. That system is still running _unattended_,
              > the reason there is a huge 40KVA UPS system with a battery room, that can
              > last for 2 days. I still give the guys a few pointers over phone ... The
              > only trouble the guys compain is when the leased lines go down, when
              > somebody breaks the roads..
              >
              > This machine was used as a mail server, SMB/NFS File server, Doubled as a
              > DHCP/DNS, and a PVM cluster server ..
              >
              > > This will be a machine as you described - no monitor, no keyboard and no
              > > mouse. I will have 1 GB of ram. And I need 1 terra byte space in Raid 5.
              > > Since this machine will be part of a production network I can't do too
              >
              > many
              >
              > Put OS onto separate HDD's itself, if you are planning for extreme low
              > downtime
              >
              > And user data into the RAID disk's/array.
              >
              > That kind of combinations depends upon what level of HA you are looking at
              >
              > For example, for my HA applications, i have a dedicated Hardware RAID U320
              > controller with 2 raid arrays A & B..
              >
              > In my case it is not the number of users, but the bandwidth that is
              > generated .. I have around 25 users only on that HA system as opposed to
              > 100 in your case.. And lot of financial data also....... That makes it HA..
              >
              > Array A contains OS and
              > Array B contains User data..
              >
              >
              > Another major suggestion that i can give you on your N/W design is that
              > most probably you will not require to have a proxy...
              >
              > You can get around using Netfilter and NAT.. That will bring down a lot of
              > your requirements... Then that machine would be a firewall machine doing
              > NAT and PAT, but again it is not advisable to use a firewalling machine as
              > your intranet server.
              >
              > Well, there are lot of points to be considered, like what exactly you are
              > trying to achieve, the permutations and combinations are huge ..
              >
              > Manu
              >
              > > experiments.
              > >
              > > I would really appreciate your help,
              > >
              > >
              > >
              > >
              > > -----Original Message-----
              > > From: Manu Abraham [mailto:manu@...]
              > > Sent: 31 December 2004 22:53
              > > To: dubailug@yahoogroups.com
              > > Subject: Re: [dubailug] Sata Drives
              > >
              > > On Fri December 31 2004 10:14 pm, cgull@... wrote:
              > > > Manu,
              > > >
              > > > Thanks for the cool info!
              > > > Can you tell us if it is possible to load this support via a module
              > > > (for
              > >
              > > It depends upon the kernel .. I had SATA support into 2.4 also One of my
              > > 2.4 machines which i use as my NFS/SAMBA fileserver, has around a total
              > > of
              > >
              > > > 200Gigs n/w storage on SATA, i have another IBM one with U320 SCA SCSI
              > >
              > > also.
              > >
              > > No problems also .. If somebody requires i can compile the kernel modules
              > > if i
              > > have that kernel version. But rather than picking up my binary modules i
              > > would really suggest guys to put in the additional pain of recompiling..
              >
              > It
              >
              > > is really worth it ..
              > >
              > > > example via insmod) rather than recompile kernel with changed options?
              > >
              > > Once compiled libata can be loaded as a module as i have loaded..
              > > You need to have to enable SCSI support also to have SATA functioning ..
              > >
              > >
              > > appletalk 33972 2
              > > ax25 58476 2
              > > ipx 26540 2
              > > ipv6 237440 18
              > > intel_mch_agp 10128 1
              > > agpgart 30888 2 intel_mch_agp
              > > bluetooth 45924 6 l2cap,hci_vhci,hci_uart,hci_usb
              > > af_packet 20872 2
              > > 3c59x 36904 0
              > > binfmt_misc 10120 1
              > > ata_piix 7684 0
              > > libata 38016 1 ata_piix,[permanent]
              > > sd_mod 20224 0
              > > scsi_mod 108876 5 sg,st,sr_mod,libata,sd_mod
              > >
              > >
              > > For older kernels, for libata support the __only__ way is to patch the
              > > kernel
              > > with the libata patch. If somebody wants to find the patch, it is
              > > maintained by Jeff Garzick. You can find the patches on kernel.org. It
              > > depends how old or what kernel it is to know what other patches if any
              > > libata requires.. I had 2.4.24 or 25... I think i had a lot of patches
              >
              > onto
              >
              > > that..
              > >
              > > I had the ck-xx patches, libata patch, lot of patches, i don't
              > > recollect.. what all.. That machine is in a corner humming away without
              > > any problems
              >
              > at
              >
              > > all as i said for more than 8 months without any interruptions.. Even
              > > during power up and downs. There is no keyboard/monitor/mouse on that..
              > > Anything goes wrong (nothing has gone wrong except power supply blew up
              > > once) somebody
              > > just presses the reset button (in emergency) else restarts through SSH..
              >
              > As
              >
              > > simple as that..
              > >
              > > I have a quite heavy data transfer on that machine GB's daily as machine
              > > does
              > > storage for some PVR jobs also.
              > >
              > > Regarding new kernels... There is no problem at all, It is like fish
              > > jumping to water.
              > >
              > > The most problematic one was the ICH5, ie, SATA on the Intel boards, but
              > > even
              > > that, the problems are no longer there with the new kernels..
              > >
              > > Kernel-2.6.9 onwards beautiful support is there, but there is exception
              > > of Seagate HDD's on SiL controllers, but only a firmware problem on the
              >
              > drives
              >
              > > as i said earlier ..
              > >
              > > Manu
              > >
              > > > cGull.
              > > >
              > > > Manu Abraham <manu@...> wrote on 31.12.2004, 18:49:14:
              > > > > On Fri December 31 2004 7:30 pm, cgull@... wrote:
              > > > > > Rajeev,
              > > > > >
              > > > > > I have tried SATA drives on HP hardware and I had some problems.
              > > > >
              > > > > Which controller was this.. It is known that Seagate HDD's and SiL
              > > > > Controllers don't get along well, but therre have been reports that
              > >
              > > HDD's
              > >
              > > > > with newer firmware does not have any problems ..
              > > > >
              > > > > > For example Knoppix and Tomsrtbt both hung on boot because of the
              > >
              > > SATA.
              > >
              > > > > libata i believe .. Try using the SATA drives in compatibity mode,
              > > > > eventhough you loose performance, but after installation of the OS,
              > > > > just rebuild the kernel with the relevant patches..
              > > > >
              > > > > > My Mandrake distro has no problems with it however.
              > > > >
              > > > > Mandrake has libata support . :-)
              > > > >
              > > > > > I searched the internet and found there were some issues with SATA.
              > > > > > To avoid this I just went for SCSI which is guaranteed to work with
              > > > > > Linux.
              > > > >
              > > > > SATA also uses the same SCSI commands under linux... In fact for a
              >
              > SATA
              >
              > > > > HDD you have the regular /dev/sdxx for SATA as compared to SCSI>
              > > > >
              > > > > > I know SATA is cheaper, but then I wasnt the one forking out the
              >
              > cash
              >
              > > > > > so I could make the choice of SCSI instead.
              > > > > > You may also have other issues such as SATA not being hot-pluggable
              > > > > > from
              > > > >
              > > > > For SCSI also, unless you have SCA SCSI you cannot Hot swap... SCA
              >
              > SCSI
              >
              > > > > is still more expensive than standard SCSI, and moreover for SCA SCSI
              > >
              > > you
              > >
              > > > > have to have the drive cage to properly Hotswap.. The only difference
              > > > > between SCA SCSI and SCSI is that the power to the SCSI disk comes
              > > > > through the SCA SCSI connector rather than the power connector.. SCA
              > >
              > > SCSI
              > >
              > > > > has no separate power connectors, but they have a controller situated
              > > > > in the drive cage to shutdown the power to the relevant drive ...
              > > > >
              > > > > > your server or disk array.
              > > > > > Tell us what hardware you plan to use (server and diskarray) and
              > > > > > the various controller cards...
              > > > >
              > > > > I have 2 x 200GB Seagate HDD's on ICH5 for quite some time.. Since
              >
              > SATA
              >
              > > > > for the ICH% stabilized only recently i had been using lot of patches
              > > > > to achieve the functionality. I never had any problem, except that i
              > > > > had a failed 160GB Western Digital SATA drive .. Has been working for
              > > > > at least 8 months i think, for me, no problems other than the
              >
              > failure..
              >
              > > > > The important point with SATA is that to have SATA support you have
              > > > > to have libata support. It depends upon which kernel you are using,
              > > > > the complexity of the procedure .. since when you have different
              > > > > patches, some patches would have lot of rejects and you would have to
              > > > > patch it manually by hand ..
              > > > >
              > > > >
              > > > > I think Brad is having 6 SATA drives, maybe he could explain his
              > > > > experience ..
              > > > >
              > > > > Manu
              > > > >
              > > > > > cGull.
              > > > > > (previously known as mattywix@...)
              > > > > >
              > > > > > Rajeev wrote on 31.12.2004, 12:47:57:
              > > > > > > Dear Friends,
              > > > > > >
              > > > > > > Happy New Year
              > > > > > >
              > > > > > > I need to build a Linux server with 1 terra byte (200gb X 6
              > > > > > > drives
              > >
              > > in
              > >
              > > > > > > raid 5) capacity. I would like to do this with SATA drives. Has
              >
              > any
              >
              > > > > > > one done this before? (SATA drives and raid 5)?
              > > > >
              > > > > No problems at all.. with recent kernels, most of the controllers are
              > > > > supported ..
              > > > >
              > > > > > > Sharing your valuable experience will be appreciated.
              > > > > > >
              > > > > > > Thanks
              > > > > > >
              > > > > > > Rajeev
              > > > > > >
              > > > > > >
              > > > > > >
              > > > > > >
              > > > > > >
              > > > > > >
              > > > > > >
              > > > > > > Yahoo! Groups Links
              > > > > > >
              > > > > > > To visit your group on the web, go to:
              > > > > > > http://groups.yahoo.com/group/dubailug/
              > > > > > >
              > > > > > > To unsubscribe from this group, send an email to:
              > > > > > > dubailug-unsubscribe@yahoogroups.com
              > > > > > >
              > > > > > > Your use of Yahoo! Groups is subject to:
              > > > > > > http://docs.yahoo.com/info/terms/
              > > > > >
              > > > > > Yahoo! Groups Links
              > > > >
              > > > > Yahoo! Groups Links
              > > > >
              > > > > To visit your group on the web, go to:
              > > > > http://groups.yahoo.com/group/dubailug/
              > > > >
              > > > > To unsubscribe from this group, send an email to:
              > > > > dubailug-unsubscribe@yahoogroups.com
              > > > >
              > > > > Your use of Yahoo! Groups is subject to:
              > > > > http://docs.yahoo.com/info/terms/
              > > >
              > > > Yahoo! Groups Links
              > >
              > > Yahoo! Groups Links
              > >
              > >
              > >
              > >
              > >
              > >
              > >
              > >
              > >
              > >
              > > Yahoo! Groups Links
              >
              > Yahoo! Groups Links
              >
              >
              >
              >
              >
              >
              >
              >
              >
              > Yahoo! Groups Links
              >
              >
              >
            • Brad Campbell
              ... Hey! Lets not get vendor specific here. There are plenty of Opteron boards that will fit the bill also! :p) ... Agreed. I have 2TB on 3 PCI cards and 10
              Message 6 of 29 , Jan 1, 2005
                Manu Abraham wrote:

                > BTW that is a 64 bit card, i think. Could not see the connectors too clearly..
                > You require a Xeon processor based machine to use that card in that case..

                Hey! Lets not get vendor specific here. There are plenty of Opteron boards that will fit the bill
                also! :p)

                > Anyway it is better to have 1TB on a 64bit/133MHz PCI bus rather than 32
                > bit/33MHz.
                >

                Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5 (which just rocks by the
                way) but a rebuild takes me nearly 9 hours. If I was on a pair of 8 port cards on PCI-X I'm sure it
                would be much quicker.

                Having said that, my read/write speeds are acceptable. Certainly not lightning quick, but given I
                use it as a server for media streaming and backups over Gig-E it does the job quite well.
                Price up 2TB of SCSI sometime and check the tradeoff.

                Probably all the details you will ever need attached below.

                Regards,
                Brad


                srv:/raid# bonnie++ -d . -u 0:0
                Using uid:0, gid:0.
                Writing with putc()...done
                Writing intelligently...done
                Rewriting...done
                Reading with getc()...done
                Reading intelligently...done
                start 'em...done...done...done...
                Create files in sequential order...done.
                Stat files in sequential order...done.
                Delete files in sequential order...done.
                Create files in random order...done.
                Stat files in random order...done.
                Delete files in random order...done.
                Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
                -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
                Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
                srv 1G 11837 37 18558 11 13577 7 19694 58 55351 20 381.1 1
                ------Sequential Create------ --------Random Create--------
                -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
                files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
                16 748 78 +++++ +++ +++++ +++ 705 82 +++++ +++ 5074 83
                srv,1G,11837,37,18558,11,13577,7,19694,58,55351,20,381.1,1,16,748,78,+++++,+++,+++++,+++,705,82,+++++,+++,5074,83

                brad@srv:~$ cat /proc/mdstat
                Personalities : [raid0] [raid5] [raid6]
                md2 : active raid5 sdl[0] sdm[2] sdk[1]
                488396800 blocks level 5, 128k chunk, algorithm 2 [3/3] [UUU]

                md0 : active raid5 sdj1[9] sda1[0] sdi1[8] sdh1[7] sdg1[6] sdf1[5] sde1[4] sdd1[3] sdc1[2] sdb1[1]
                2206003968 blocks level 5, 128k chunk, algorithm 0 [10/10] [UUUUUUUUUU]

                unused devices: <none>

                srv:/home/brad# mdadm -D /dev/md0
                /dev/md0:
                Version : 00.90.01
                Creation Time : Sun May 2 18:02:14 2004
                Raid Level : raid5
                Array Size : 2206003968 (2103.81 GiB 2258.95 GB)
                Device Size : 245111552 (233.76 GiB 250.99 GB)
                Raid Devices : 10
                Total Devices : 10
                Preferred Minor : 0
                Persistence : Superblock is persistent

                Update Time : Sun Jan 2 10:00:08 2005
                State : clean
                Active Devices : 10
                Working Devices : 10
                Failed Devices : 0
                Spare Devices : 0

                Layout : left-asymmetric
                Chunk Size : 128K

                UUID : 05cc3f43:de1ecfa4:83a51293:78015f1e
                Events : 0.1052890

                Number Major Minor RaidDevice State
                0 8 1 0 active sync /dev/devfs/scsi/host0/bus0/target0/lun0/part1
                1 8 17 1 active sync /dev/devfs/scsi/host1/bus0/target0/lun0/part1
                2 8 33 2 active sync /dev/devfs/scsi/host2/bus0/target0/lun0/part1
                3 8 49 3 active sync /dev/devfs/scsi/host3/bus0/target0/lun0/part1
                4 8 65 4 active sync /dev/devfs/scsi/host4/bus0/target0/lun0/part1
                5 8 81 5 active sync /dev/devfs/scsi/host5/bus0/target0/lun0/part1
                6 8 97 6 active sync /dev/devfs/scsi/host6/bus0/target0/lun0/part1
                7 8 113 7 active sync /dev/devfs/scsi/host7/bus0/target0/lun0/part1
                8 8 129 8 active sync /dev/sdi1
                9 8 145 9 active sync /dev/sdj1


                brad@srv:~$ lspci
                0000:00:00.0 Host bridge: VIA Technologies, Inc. VT8377 [KT400/KT600 AGP] Host Bridge (rev 80)
                0000:00:01.0 PCI bridge: VIA Technologies, Inc. VT8237 PCI Bridge
                0000:00:09.0 Ethernet controller: 3Com Corporation 3c940 10/100/1000Base-T [Marvell] (rev 12)
                0000:00:0a.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
                0000:00:0b.0 Unknown mass storage controller: Promise Technology, Inc. PDC20318 (SATA150 TX4) (rev 02)
                0000:00:0c.0 FireWire (IEEE 1394): VIA Technologies, Inc. IEEE 1394 Host Controller (rev 46)
                0000:00:0d.0 Ethernet controller: Marvell Technology Group Ltd. Yukon Gigabit Ethernet
                10/100/1000Base-T Adapter (rev 12)
                0000:00:0e.0 Unknown mass storage controller: Promise Technology, Inc. PDC20318 (SATA150 TX4) (rev 02)
                0000:00:0f.0 RAID bus controller: VIA Technologies, Inc. VIA VT6420 SATA RAID Controller (rev 80)
                0000:00:0f.1 IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT823x/A/C PIPC Bus
                Master IDE (rev 06)
                0000:00:10.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 81)
                0000:00:10.1 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 81)
                0000:00:10.2 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 81)
                0000:00:10.3 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 81)
                0000:00:10.4 USB Controller: VIA Technologies, Inc. USB 2.0 (rev 86)
                0000:00:11.0 ISA bridge: VIA Technologies, Inc. VT8237 ISA bridge [K8T800 South]
                0000:00:13.0 Unknown mass storage controller: Promise Technology, Inc. PDC20318 (SATA150 TX4) (rev 02)
                0000:01:00.0 VGA compatible controller: Silicon Integrated Systems [SiS] 315PRO PCI/AGP VGA Display
                Adapter

                brad@srv:~$ getsmart

                Disk - /dev/sda
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y61DXT7E
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:43 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 106) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 176 176 063 Pre-fail Always - 23157
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 173
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 247 187 Pre-fail Always - 49463
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 529h+22m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 222
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 39
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 21585
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 4
                202 TA_Increase_Count 0x000a 253 250 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 16
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 192 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5612 -
                # 2 Extended offline Completed without error 00% 5611 -
                # 3 Short offline Completed without error 00% 5589 -
                # 4 Short offline Completed without error 00% 5567 -
                # 5 Short offline Completed without error 00% 5552 -
                # 6 Short offline Completed without error 00% 5522 -
                # 7 Short offline Completed without error 00% 5500 -
                # 8 Short offline Completed without error 00% 5478 -
                # 9 Short offline Completed without error 00% 5455 -
                #10 Extended offline Completed without error 00% 5455 -
                #11 Short offline Completed without error 00% 5433 -
                #12 Short offline Completed without error 00% 5411 -
                #13 Short offline Completed without error 00% 5388 -
                #14 Short offline Completed without error 00% 5366 -
                #15 Short offline Completed without error 00% 5343 -
                #16 Short offline Completed without error 00% 5321 -
                #17 Short offline Completed without error 00% 5299 -
                #18 Extended offline Completed without error 00% 5298 -
                #19 Short offline Completed without error 00% 5276 -
                #20 Short offline Completed without error 00% 5254 -
                #21 Short offline Completed without error 00% 5232 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdb
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y61DY0FE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:43 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 106) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 176 176 063 Pre-fail Always - 17756
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 156
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 48047
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 490h+22m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 203
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 39
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 8761
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 5
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 0
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5575 -
                # 2 Extended offline Completed without error 00% 5575 -
                # 3 Short offline Completed without error 00% 5553 -
                # 4 Short offline Completed without error 00% 5530 -
                # 5 Short offline Completed without error 00% 5515 -
                # 6 Short offline Completed without error 00% 5486 -
                # 7 Short offline Completed without error 00% 5463 -
                # 8 Short offline Completed without error 00% 5441 -
                # 9 Short offline Completed without error 00% 5419 -
                #10 Extended offline Completed without error 00% 5418 -
                #11 Short offline Completed without error 00% 5396 -
                #12 Short offline Completed without error 00% 5374 -
                #13 Short offline Completed without error 00% 5352 -
                #14 Short offline Completed without error 00% 5329 -
                #15 Short offline Completed without error 00% 5307 -
                #16 Short offline Completed without error 00% 5285 -
                #17 Short offline Completed without error 00% 5262 -
                #18 Extended offline Completed without error 00% 5262 -
                #19 Short offline Completed without error 00% 5240 -
                #20 Short offline Completed without error 00% 5217 -
                #21 Short offline Completed without error 00% 5195 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdc
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y61DXB7E
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:44 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 106) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 175 175 063 Pre-fail Always - 18371
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 157
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 48433
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 501h+44m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 202
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 39
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 12683
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 6
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 6
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5586 -
                # 2 Extended offline Completed without error 00% 5585 -
                # 3 Short offline Completed without error 00% 5563 -
                # 4 Short offline Completed without error 00% 5541 -
                # 5 Short offline Completed without error 00% 5526 -
                # 6 Short offline Completed without error 00% 5496 -
                # 7 Short offline Completed without error 00% 5474 -
                # 8 Short offline Completed without error 00% 5452 -
                # 9 Short offline Completed without error 00% 5429 -
                #10 Extended offline Completed without error 00% 5429 -
                #11 Short offline Completed without error 00% 5407 -
                #12 Short offline Completed without error 00% 5385 -
                #13 Short offline Completed without error 00% 5362 -
                #14 Short offline Completed without error 00% 5340 -
                #15 Short offline Completed without error 00% 5318 -
                #16 Short offline Completed without error 00% 5295 -
                #17 Short offline Completed without error 00% 5273 -
                #18 Extended offline Completed without error 00% 5273 -
                #19 Short offline Completed without error 00% 5250 -
                #20 Short offline Completed without error 00% 5228 -
                #21 Short offline Completed without error 00% 5206 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdd
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y61DXBFE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:44 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 107) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 174 174 063 Pre-fail Always - 17961
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 155
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 49516
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 488h+24m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 200
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 37
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 22874
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 3
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 23
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5573 -
                # 2 Extended offline Completed without error 00% 5573 -
                # 3 Short offline Completed without error 00% 5551 -
                # 4 Short offline Completed without error 00% 5529 -
                # 5 Short offline Completed without error 00% 5513 -
                # 6 Short offline Completed without error 00% 5484 -
                # 7 Short offline Completed without error 00% 5462 -
                # 8 Short offline Completed without error 00% 5439 -
                # 9 Short offline Completed without error 00% 5417 -
                #10 Extended offline Completed without error 00% 5416 -
                #11 Short offline Completed without error 00% 5394 -
                #12 Short offline Completed without error 00% 5372 -
                #13 Short offline Completed without error 00% 5350 -
                #14 Short offline Completed without error 00% 5327 -
                #15 Short offline Completed without error 00% 5305 -
                #16 Short offline Completed without error 00% 5283 -
                #17 Short offline Completed without error 00% 5260 -
                #18 Extended offline Completed without error 00% 5260 -
                #19 Short offline Completed without error 00% 5238 -
                #20 Short offline Completed without error 00% 5216 -
                #21 Short offline Completed without error 00% 5193 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sde
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y61DX8QE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:44 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 107) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 176 176 063 Pre-fail Always - 17425
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 155
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 49275
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 488h+23m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 201
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 37
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 9618
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 1
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 13
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5573 -
                # 2 Extended offline Completed without error 00% 5573 -
                # 3 Short offline Completed without error 00% 5551 -
                # 4 Short offline Completed without error 00% 5529 -
                # 5 Short offline Completed without error 00% 5513 -
                # 6 Short offline Completed without error 00% 5484 -
                # 7 Short offline Completed without error 00% 5461 -
                # 8 Short offline Completed without error 00% 5439 -
                # 9 Short offline Completed without error 00% 5417 -
                #10 Extended offline Completed without error 00% 5416 -
                #11 Short offline Completed without error 00% 5394 -
                #12 Short offline Completed without error 00% 5372 -
                #13 Short offline Completed without error 00% 5350 -
                #14 Short offline Completed without error 00% 5327 -
                #15 Short offline Completed without error 00% 5305 -
                #16 Short offline Completed without error 00% 5283 -
                #17 Short offline Completed without error 00% 5260 -
                #18 Extended offline Completed without error 00% 5260 -
                #19 Short offline Completed without error 00% 5238 -
                #20 Short offline Completed without error 00% 5216 -
                #21 Short offline Completed without error 00% 5193 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdf
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y627E3HE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:44 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 106) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 175 174 063 Pre-fail Always - 17356
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 154
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 47621
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 484h+34m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 201
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 38
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 2657
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 4
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 0
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 194 190 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5570 -
                # 2 Extended offline Completed without error 00% 5569 -
                # 3 Short offline Completed without error 00% 5547 -
                # 4 Short offline Completed without error 00% 5525 -
                # 5 Short offline Completed without error 00% 5510 -
                # 6 Short offline Completed without error 00% 5480 -
                # 7 Short offline Completed without error 00% 5458 -
                # 8 Short offline Completed without error 00% 5436 -
                # 9 Short offline Completed without error 00% 5413 -
                #10 Extended offline Completed without error 00% 5413 -
                #11 Short offline Completed without error 00% 5391 -
                #12 Short offline Completed without error 00% 5369 -
                #13 Short offline Completed without error 00% 5346 -
                #14 Short offline Completed without error 00% 5324 -
                #15 Short offline Completed without error 00% 5301 -
                #16 Short offline Completed without error 00% 5279 -
                #17 Short offline Completed without error 00% 5257 -
                #18 Extended offline Completed without error 00% 5256 -
                #19 Short offline Completed without error 00% 5234 -
                #20 Short offline Completed without error 00% 5212 -
                #21 Short offline Completed without error 00% 5190 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdg
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y628S4EE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:45 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 107) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 175 175 063 Pre-fail Always - 17418
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 154
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 47749
                9 Power_On_Minutes 0x0032 236 236 000 Old_age Always - 484h+39m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 200
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 38
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 3413
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 1
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 0
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 197 188 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5570 -
                # 2 Extended offline Completed without error 00% 5569 -
                # 3 Short offline Completed without error 00% 5547 -
                # 4 Short offline Completed without error 00% 5525 -
                # 5 Short offline Completed without error 00% 5510 -
                # 6 Short offline Completed without error 00% 5480 -
                # 7 Short offline Completed without error 00% 5458 -
                # 8 Short offline Completed without error 00% 5436 -
                # 9 Short offline Completed without error 00% 5413 -
                #10 Extended offline Completed without error 00% 5413 -
                #11 Short offline Completed without error 00% 5391 -
                #12 Short offline Completed without error 00% 5369 -
                #13 Short offline Completed without error 00% 5346 -
                #14 Short offline Completed without error 00% 5324 -
                #15 Short offline Completed without error 00% 5302 -
                #16 Short offline Completed without error 00% 5279 -
                #17 Short offline Completed without error 00% 5257 -
                #18 Extended offline Completed without error 00% 5256 -
                #19 Short offline Completed without error 00% 5234 -
                #20 Short offline Completed without error 00% 5212 -
                #21 Short offline Completed without error 00% 5190 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdh
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y625DKCE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:45 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 107) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 175 175 063 Pre-fail Always - 18693
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 58
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 47887
                9 Power_On_Minutes 0x0032 237 237 000 Old_age Always - 206h+53m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 81
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 40
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 5485
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 3
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 1
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_age Always - 0
                208 Spin_Buzz 0x002a 253 252 000 Old_age Always - 0
                209 Offline_Seek_Performnce 0x0024 196 189 000 Old_age Offline - 0
                99 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                100 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0
                101 Unknown_Attribute 0x0004 253 253 000 Old_age Offline - 0

                SMART Error Log Version: 1
                No Errors Logged

                SMART Self-test log structure revision number 1
                Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
                # 1 Short offline Completed without error 00% 5309 -
                # 2 Extended offline Completed without error 00% 5309 -
                # 3 Short offline Completed without error 00% 5287 -
                # 4 Short offline Completed without error 00% 5265 -
                # 5 Short offline Completed without error 00% 5249 -
                # 6 Short offline Completed without error 00% 5220 -
                # 7 Short offline Completed without error 00% 5198 -
                # 8 Short offline Completed without error 00% 5175 -
                # 9 Short offline Completed without error 00% 5153 -
                #10 Extended offline Completed without error 00% 5153 -
                #11 Short offline Completed without error 00% 5131 -
                #12 Short offline Completed without error 00% 5108 -
                #13 Short offline Completed without error 00% 5086 -
                #14 Short offline Completed without error 00% 5063 -
                #15 Short offline Completed without error 00% 5041 -
                #16 Short offline Completed without error 00% 5019 -
                #17 Short offline Completed without error 00% 4996 -
                #18 Extended offline Completed without error 00% 4996 -
                #19 Short offline Completed without error 00% 4974 -
                #20 Short offline Completed without error 00% 4952 -
                #21 Short offline Completed without error 00% 4929 -

                SMART Selective self-test log data structure revision number 1
                SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                1 0 0 Not_testing
                2 0 0 Not_testing
                3 0 0 Not_testing
                4 0 0 Not_testing
                5 0 0 Not_testing
                Selective self-test flags (0x0):
                After scanning selected spans, do NOT read-scan remainder of disk.
                If Selective self-test is pending on power-up, resume after 0 minute delay.

                Disk - /dev/sdi
                smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                Home page is http://smartmontools.sourceforge.net/

                === START OF INFORMATION SECTION ===
                Device Model: Maxtor 7Y250M0
                Serial Number: Y625W7ZE
                Firmware Version: YAR51EW0
                Device is: In smartctl database [for details use: -P show]
                ATA Version is: 7
                ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                Local Time is: Sun Jan 2 09:44:45 2005 GST
                SMART support is: Available - device has SMART capability.
                SMART support is: Enabled

                === START OF READ SMART DATA SECTION ===
                SMART overall-health self-assessment test result: PASSED

                General SMART Values:
                Offline data collection status: (0x82) Offline data collection activity
                was completed without error.
                Auto Offline Data Collection: Enabled.
                Self-test execution status: ( 0) The previous self-test routine completed
                without error or no self-test has ever
                been run.
                Total time to complete Offline
                data collection: ( 363) seconds.
                Offline data collection
                capabilities: (0x5b) SMART execute Offline immediate.
                Auto Offline data collection on/off support.
                Suspend Offline collection upon new
                command.
                Offline surface scan supported.
                Self-test supported.
                No Conveyance Self-test supported.
                Selective Self-test supported.
                SMART capabilities: (0x0003) Saves SMART data before entering
                power-saving mode.
                Supports SMART auto save timer.
                Error logging capability: (0x01) Error logging supported.
                No General Purpose Logging support.
                Short self-test routine
                recommended polling time: ( 2) minutes.
                Extended self-test routine
                recommended polling time: ( 106) minutes.

                SMART Attributes Data Structure revision number: 16
                Vendor Specific SMART Attributes with Thresholds:
                ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
                3 Spin_Up_Time 0x0027 174 174 063 Pre-fail Always - 18865
                4 Start_Stop_Count 0x0032 253 253 000 Old_age Always - 58
                5 Reallocated_Sector_Ct 0x0033 253 253 063 Pre-fail Always - 0
                6 Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline - 0
                7 Seek_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                8 Seek_Time_Performance 0x0027 251 246 187 Pre-fail Always - 48669
                9 Power_On_Minutes 0x0032 237 237 000 Old_age Always - 206h+50m
                10 Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always - 0
                11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail Always - 0
                12 Power_Cycle_Count 0x0032 253 253 000 Old_age Always - 81
                192 Power-Off_Retract_Count 0x0032 253 253 000 Old_age Always - 0
                193 Load_Cycle_Count 0x0032 253 253 000 Old_age Always - 0
                194 Temperature_Celsius 0x0032 253 253 000 Old_age Always - 39
                195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age Always - 4198
                196 Reallocated_Event_Count 0x0008 253 253 000 Old_age Offline - 0
                197 Current_Pending_Sector 0x0008 253 253 000 Old_age Offline - 0
                198 Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline - 0
                199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age Offline - 0
                200 Multi_Zone_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                201 Soft_Read_Error_Rate 0x000a 253 252 000 Old_age Always - 0
                202 TA_Increase_Count 0x000a 253 252 000 Old_age Always - 0
                203 Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always - 0
                204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                205 Shock_Rate_Write_Opern 0x000a 253 252 000 Old_age Always - 0
                207 Spin_High_Current 0x002a 253 252 000 Old_a<br/><br/>(Message over 64 KB, truncated)
              • Manu Abraham
                ... The reason i got vendor specific was only one... If you go for an Opteron processor, stick to AMD PCI implementations rather than VIA implementations .. My
                Message 7 of 29 , Jan 1, 2005
                  On Sun January 2 2005 10:04 am, Brad Campbell wrote:
                  > Manu Abraham wrote:
                  > > BTW that is a 64 bit card, i think. Could not see the connectors too
                  > > clearly.. You require a Xeon processor based machine to use that card in
                  > > that case..
                  >
                  > Hey! Lets not get vendor specific here. There are plenty of Opteron boards
                  > that will fit the bill also! :p)
                  >
                  The reason i got vendor specific was only one... If you go for an Opteron
                  processor, stick to AMD PCI implementations rather than VIA
                  implementations ..

                  My story goes like this ...

                  I used to like AMD chips a lot, but proper motherboards have prevented a lot
                  of people from diving into that.. and that is the main reason why Intel
                  thrives. The processor, North bridge/Southbridge implementation on the Intel
                  side is quite tight, but when it comes to AMD, eventhough the processor is
                  very superior, the motherboard implementation on the AMD side is quite loose.
                  Exception for the motherboards from AMD. But again i would ike to a have a
                  choice rather than somebody say, if you have an AMD chip, use this
                  motherboard.. There are more things that i would be looking into rather than
                  just the vendor alone, like number of slots, type of slots etc ...

                  Well, that's for me.. not for everybody..

                  I do a lot of work on the media, and for most of the work the VIA chipsets are
                  hopeless, due to their cheap and bad implementation of the PCI bus timings..

                  I get a lot of problems with VIA chipsets on the Twinhan and the Kfir.. So
                  whatever it might be, VIA chipsets are a strict NO... NO... for me at least
                  and for anybody who uses cards that heavily depend upon PCI bus timings,
                  Examples are industry standard MPEG encoders, cheap
                  Satellite/Cable/Terrestrial cards .. and the list goes on ..

                  > > Anyway it is better to have 1TB on a 64bit/133MHz PCI bus rather than 32
                  > > bit/33MHz.
                  >
                  > Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5
                  > (which just rocks by the way) but a rebuild takes me nearly 9 hours. If I
                  > was on a pair of 8 port cards on PCI-X I'm sure it would be much quicker.
                  >
                  What if something bad again happened during the rebuild ? It is really
                  dangerous to have very long periods of rebuilds .. For the same reason only
                  Adaptec and other major manufacturers have battery backed up Hardware RAID
                  cards ..

                  > Having said that, my read/write speeds are acceptable. Certainly not
                  > lightning quick, but given I use it as a server for media streaming and
                  > backups over Gig-E it does the job quite well. Price up 2TB of SCSI
                  > sometime and check the tradeoff.

                  Parallel SCSI is dead, the new implementations of SCSI are serial in nature,
                  like iSCSI and so on.. Only for backward compatibility reasons parallel SCSI
                  do exist for the time being ..

                  Manu
                  >
                  > Probably all the details you will ever need attached below.
                  >
                  > Regards,
                  > Brad
                  >
                  >
                  > srv:/raid# bonnie++ -d . -u 0:0
                  > Using uid:0, gid:0.
                  > Writing with putc()...done
                  > Writing intelligently...done
                  > Rewriting...done
                  > Reading with getc()...done
                  > Reading intelligently...done
                  > start 'em...done...done...done...
                  > Create files in sequential order...done.
                  > Stat files in sequential order...done.
                  > Delete files in sequential order...done.
                  > Create files in random order...done.
                  > Stat files in random order...done.
                  > Delete files in random order...done.
                  > Version 1.03 ------Sequential Output------ --Sequential Input-
                  > --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
                  > Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec
                  > %CP srv 1G 11837 37 18558 11 13577 7 19694 58 55351 20
                  > 381.1 1 ------Sequential Create------ --------Random Create--------
                  > -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP
                  > /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 748 78 +++++ +++
                  > +++++ +++ 705 82 +++++ +++ 5074 83
                  > srv,1G,11837,37,18558,11,13577,7,19694,58,55351,20,381.1,1,16,748,78,+++++,
                  >+++,+++++,+++,705,82,+++++,+++,5074,83
                  >
                  > brad@srv:~$ cat /proc/mdstat
                  > Personalities : [raid0] [raid5] [raid6]
                  > md2 : active raid5 sdl[0] sdm[2] sdk[1]
                  > 488396800 blocks level 5, 128k chunk, algorithm 2 [3/3] [UUU]
                  >
                  > md0 : active raid5 sdj1[9] sda1[0] sdi1[8] sdh1[7] sdg1[6] sdf1[5] sde1[4]
                  > sdd1[3] sdc1[2] sdb1[1] 2206003968 blocks level 5, 128k chunk, algorithm 0
                  > [10/10] [UUUUUUUUUU]
                  >
                  > unused devices: <none>
                  >
                  > srv:/home/brad# mdadm -D /dev/md0
                  > /dev/md0:
                  > Version : 00.90.01
                  > Creation Time : Sun May 2 18:02:14 2004
                  > Raid Level : raid5
                  > Array Size : 2206003968 (2103.81 GiB 2258.95 GB)
                  > Device Size : 245111552 (233.76 GiB 250.99 GB)
                  > Raid Devices : 10
                  > Total Devices : 10
                  > Preferred Minor : 0
                  > Persistence : Superblock is persistent
                  >
                  > Update Time : Sun Jan 2 10:00:08 2005
                  > State : clean
                  > Active Devices : 10
                  > Working Devices : 10
                  > Failed Devices : 0
                  > Spare Devices : 0
                  >
                  > Layout : left-asymmetric
                  > Chunk Size : 128K
                  >
                  > UUID : 05cc3f43:de1ecfa4:83a51293:78015f1e
                  > Events : 0.1052890
                  >
                  > Number Major Minor RaidDevice State
                  > 0 8 1 0 active sync
                  > /dev/devfs/scsi/host0/bus0/target0/lun0/part1 1 8 17 1
                  > active sync /dev/devfs/scsi/host1/bus0/target0/lun0/part1 2 8
                  > 33 2 active sync
                  > /dev/devfs/scsi/host2/bus0/target0/lun0/part1 3 8 49 3
                  > active sync /dev/devfs/scsi/host3/bus0/target0/lun0/part1 4 8
                  > 65 4 active sync
                  > /dev/devfs/scsi/host4/bus0/target0/lun0/part1 5 8 81 5
                  > active sync /dev/devfs/scsi/host5/bus0/target0/lun0/part1 6 8
                  > 97 6 active sync
                  > /dev/devfs/scsi/host6/bus0/target0/lun0/part1 7 8 113 7
                  > active sync /dev/devfs/scsi/host7/bus0/target0/lun0/part1 8 8
                  > 129 8 active sync /dev/sdi1
                  > 9 8 145 9 active sync /dev/sdj1
                  >
                  >
                  > brad@srv:~$ lspci
                  > 0000:00:00.0 Host bridge: VIA Technologies, Inc. VT8377 [KT400/KT600 AGP]
                  > Host Bridge (rev 80) 0000:00:01.0 PCI bridge: VIA Technologies, Inc. VT8237
                  > PCI Bridge
                  > 0000:00:09.0 Ethernet controller: 3Com Corporation 3c940 10/100/1000Base-T
                  > [Marvell] (rev 12) 0000:00:0a.0 Ethernet controller: Realtek Semiconductor
                  > Co., Ltd. RTL-8139/8139C/8139C+ (rev 10) 0000:00:0b.0 Unknown mass storage
                  > controller: Promise Technology, Inc. PDC20318 (SATA150 TX4) (rev 02)
                  > 0000:00:0c.0 FireWire (IEEE 1394): VIA Technologies, Inc. IEEE 1394 Host
                  > Controller (rev 46) 0000:00:0d.0 Ethernet controller: Marvell Technology
                  > Group Ltd. Yukon Gigabit Ethernet 10/100/1000Base-T Adapter (rev 12)
                  > 0000:00:0e.0 Unknown mass storage controller: Promise Technology, Inc.
                  > PDC20318 (SATA150 TX4) (rev 02) 0000:00:0f.0 RAID bus controller: VIA
                  > Technologies, Inc. VIA VT6420 SATA RAID Controller (rev 80) 0000:00:0f.1
                  > IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT823x/A/C
                  > PIPC Bus Master IDE (rev 06)
                  > 0000:00:10.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1
                  > Controller (rev 81) 0000:00:10.1 USB Controller: VIA Technologies, Inc.
                  > VT82xxxxx UHCI USB 1.1 Controller (rev 81) 0000:00:10.2 USB Controller: VIA
                  > Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 81) 0000:00:10.3
                  > USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller
                  > (rev 81) 0000:00:10.4 USB Controller: VIA Technologies, Inc. USB 2.0 (rev
                  > 86) 0000:00:11.0 ISA bridge: VIA Technologies, Inc. VT8237 ISA bridge
                  > [K8T800 South] 0000:00:13.0 Unknown mass storage controller: Promise
                  > Technology, Inc. PDC20318 (SATA150 TX4) (rev 02) 0000:01:00.0 VGA
                  > compatible controller: Silicon Integrated Systems [SiS] 315PRO PCI/AGP VGA
                  > Display Adapter
                  >
                  > brad@srv:~$ getsmart
                  >
                  > Disk - /dev/sda
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y61DXT7E
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:43 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 106) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 176 176 063
                  > Pre-fail Always - 23157 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 173 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 247 187
                  > Pre-fail Always - 49463 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 529h+22m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 222 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 39 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 21585 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 4 202 TA_Increase_Count
                  > 0x000a 253 250 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 16 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252
                  > 000 Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 192 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5612 - # 2 Extended offline Completed without error
                  > 00% 5611 - # 3 Short offline Completed without error
                  > 00% 5589 - # 4 Short offline Completed without
                  > error 00% 5567 - # 5 Short offline Completed
                  > without error 00% 5552 - # 6 Short offline
                  > Completed without error 00% 5522 - # 7 Short offline
                  > Completed without error 00% 5500 - # 8 Short offline
                  > Completed without error 00% 5478 - # 9 Short
                  > offline Completed without error 00% 5455 - #10
                  > Extended offline Completed without error 00% 5455 -
                  > #11 Short offline Completed without error 00% 5433
                  > - #12 Short offline Completed without error 00% 5411
                  > - #13 Short offline Completed without error 00% 5388
                  > - #14 Short offline Completed without error 00%
                  > 5366 - #15 Short offline Completed without error 00%
                  > 5343 - #16 Short offline Completed without error
                  > 00% 5321 - #17 Short offline Completed without error
                  > 00% 5299 - #18 Extended offline Completed without error
                  > 00% 5298 - #19 Short offline Completed without
                  > error 00% 5276 - #20 Short offline Completed
                  > without error 00% 5254 - #21 Short offline
                  > Completed without error 00% 5232 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdb
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y61DY0FE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:43 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 106) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 176 176 063
                  > Pre-fail Always - 17756 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 156 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 48047 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 490h+22m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 203 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 39 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 8761 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 5 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 0 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252 000
                  > Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5575 - # 2 Extended offline Completed without error
                  > 00% 5575 - # 3 Short offline Completed without error
                  > 00% 5553 - # 4 Short offline Completed without
                  > error 00% 5530 - # 5 Short offline Completed
                  > without error 00% 5515 - # 6 Short offline
                  > Completed without error 00% 5486 - # 7 Short offline
                  > Completed without error 00% 5463 - # 8 Short offline
                  > Completed without error 00% 5441 - # 9 Short
                  > offline Completed without error 00% 5419 - #10
                  > Extended offline Completed without error 00% 5418 -
                  > #11 Short offline Completed without error 00% 5396
                  > - #12 Short offline Completed without error 00% 5374
                  > - #13 Short offline Completed without error 00% 5352
                  > - #14 Short offline Completed without error 00%
                  > 5329 - #15 Short offline Completed without error 00%
                  > 5307 - #16 Short offline Completed without error
                  > 00% 5285 - #17 Short offline Completed without error
                  > 00% 5262 - #18 Extended offline Completed without error
                  > 00% 5262 - #19 Short offline Completed without
                  > error 00% 5240 - #20 Short offline Completed
                  > without error 00% 5217 - #21 Short offline
                  > Completed without error 00% 5195 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdc
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y61DXB7E
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:44 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 106) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 175 175 063
                  > Pre-fail Always - 18371 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 157 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 48433 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 501h+44m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 202 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 39 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 12683 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 6 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 6 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252 000
                  > Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5586 - # 2 Extended offline Completed without error
                  > 00% 5585 - # 3 Short offline Completed without error
                  > 00% 5563 - # 4 Short offline Completed without
                  > error 00% 5541 - # 5 Short offline Completed
                  > without error 00% 5526 - # 6 Short offline
                  > Completed without error 00% 5496 - # 7 Short offline
                  > Completed without error 00% 5474 - # 8 Short offline
                  > Completed without error 00% 5452 - # 9 Short
                  > offline Completed without error 00% 5429 - #10
                  > Extended offline Completed without error 00% 5429 -
                  > #11 Short offline Completed without error 00% 5407
                  > - #12 Short offline Completed without error 00% 5385
                  > - #13 Short offline Completed without error 00% 5362
                  > - #14 Short offline Completed without error 00%
                  > 5340 - #15 Short offline Completed without error 00%
                  > 5318 - #16 Short offline Completed without error
                  > 00% 5295 - #17 Short offline Completed without error
                  > 00% 5273 - #18 Extended offline Completed without error
                  > 00% 5273 - #19 Short offline Completed without
                  > error 00% 5250 - #20 Short offline Completed
                  > without error 00% 5228 - #21 Short offline
                  > Completed without error 00% 5206 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdd
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y61DXBFE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:44 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 107) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 174 174 063
                  > Pre-fail Always - 17961 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 155 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 49516 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 488h+24m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 200 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 37 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 22874 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 3 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 23 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252
                  > 000 Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5573 - # 2 Extended offline Completed without error
                  > 00% 5573 - # 3 Short offline Completed without error
                  > 00% 5551 - # 4 Short offline Completed without
                  > error 00% 5529 - # 5 Short offline Completed
                  > without error 00% 5513 - # 6 Short offline
                  > Completed without error 00% 5484 - # 7 Short offline
                  > Completed without error 00% 5462 - # 8 Short offline
                  > Completed without error 00% 5439 - # 9 Short
                  > offline Completed without error 00% 5417 - #10
                  > Extended offline Completed without error 00% 5416 -
                  > #11 Short offline Completed without error 00% 5394
                  > - #12 Short offline Completed without error 00% 5372
                  > - #13 Short offline Completed without error 00% 5350
                  > - #14 Short offline Completed without error 00%
                  > 5327 - #15 Short offline Completed without error 00%
                  > 5305 - #16 Short offline Completed without error
                  > 00% 5283 - #17 Short offline Completed without error
                  > 00% 5260 - #18 Extended offline Completed without error
                  > 00% 5260 - #19 Short offline Completed without
                  > error 00% 5238 - #20 Short offline Completed
                  > without error 00% 5216 - #21 Short offline
                  > Completed without error 00% 5193 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sde
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y61DX8QE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:44 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 107) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 176 176 063
                  > Pre-fail Always - 17425 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 155 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 49275 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 488h+23m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 201 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 37 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 9618 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 1 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 13 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252
                  > 000 Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 190 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5573 - # 2 Extended offline Completed without error
                  > 00% 5573 - # 3 Short offline Completed without error
                  > 00% 5551 - # 4 Short offline Completed without
                  > error 00% 5529 - # 5 Short offline Completed
                  > without error 00% 5513 - # 6 Short offline
                  > Completed without error 00% 5484 - # 7 Short offline
                  > Completed without error 00% 5461 - # 8 Short offline
                  > Completed without error 00% 5439 - # 9 Short
                  > offline Completed without error 00% 5417 - #10
                  > Extended offline Completed without error 00% 5416 -
                  > #11 Short offline Completed without error 00% 5394
                  > - #12 Short offline Completed without error 00% 5372
                  > - #13 Short offline Completed without error 00% 5350
                  > - #14 Short offline Completed without error 00%
                  > 5327 - #15 Short offline Completed without error 00%
                  > 5305 - #16 Short offline Completed without error
                  > 00% 5283 - #17 Short offline Completed without error
                  > 00% 5260 - #18 Extended offline Completed without error
                  > 00% 5260 - #19 Short offline Completed without
                  > error 00% 5238 - #20 Short offline Completed
                  > without error 00% 5216 - #21 Short offline
                  > Completed without error 00% 5193 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdf
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y627E3HE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:44 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 106) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 175 174 063
                  > Pre-fail Always - 17356 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 154 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 47621 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 484h+34m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 201 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 38 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 2657 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 4 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 0 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252 000
                  > Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 194 190 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5570 - # 2 Extended offline Completed without error
                  > 00% 5569 - # 3 Short offline Completed without error
                  > 00% 5547 - # 4 Short offline Completed without
                  > error 00% 5525 - # 5 Short offline Completed
                  > without error 00% 5510 - # 6 Short offline
                  > Completed without error 00% 5480 - # 7 Short offline
                  > Completed without error 00% 5458 - # 8 Short offline
                  > Completed without error 00% 5436 - # 9 Short
                  > offline Completed without error 00% 5413 - #10
                  > Extended offline Completed without error 00% 5413 -
                  > #11 Short offline Completed without error 00% 5391
                  > - #12 Short offline Completed without error 00% 5369
                  > - #13 Short offline Completed without error 00% 5346
                  > - #14 Short offline Completed without error 00%
                  > 5324 - #15 Short offline Completed without error 00%
                  > 5301 - #16 Short offline Completed without error
                  > 00% 5279 - #17 Short offline Completed without error
                  > 00% 5257 - #18 Extended offline Completed without error
                  > 00% 5256 - #19 Short offline Completed without
                  > error 00% 5234 - #20 Short offline Completed
                  > without error 00% 5212 - #21 Short offline
                  > Completed without error 00% 5190 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdg
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y628S4EE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:45 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 107) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 175 175 063
                  > Pre-fail Always - 17418 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 154 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 47749 9 Power_On_Minutes 0x0032
                  > 236 236 000 Old_age Always - 484h+39m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 200 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 38 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 3413 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 1 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 0 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252 000
                  > Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 197 188 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5570 - # 2 Extended offline Completed without error
                  > 00% 5569 - # 3 Short offline Completed without error
                  > 00% 5547 - # 4 Short offline Completed without
                  > error 00% 5525 - # 5 Short offline Completed
                  > without error 00% 5510 - # 6 Short offline
                  > Completed without error 00% 5480 - # 7 Short offline
                  > Completed without error 00% 5458 - # 8 Short offline
                  > Completed without error 00% 5436 - # 9 Short
                  > offline Completed without error 00% 5413 - #10
                  > Extended offline Completed without error 00% 5413 -
                  > #11 Short offline Completed without error 00% 5391
                  > - #12 Short offline Completed without error 00% 5369
                  > - #13 Short offline Completed without error 00% 5346
                  > - #14 Short offline Completed without error 00%
                  > 5324 - #15 Short offline Completed without error 00%
                  > 5302 - #16 Short offline Completed without error
                  > 00% 5279 - #17 Short offline Completed without error
                  > 00% 5257 - #18 Extended offline Completed without error
                  > 00% 5256 - #19 Short offline Completed without
                  > error 00% 5234 - #20 Short offline Completed
                  > without error 00% 5212 - #21 Short offline
                  > Completed without error 00% 5190 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdh
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y625DKCE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:45 2005 GST
                  > SMART support is: Available - device has SMART capability.
                  > SMART support is: Enabled
                  >
                  > === START OF READ SMART DATA SECTION ===
                  > SMART overall-health self-assessment test result: PASSED
                  >
                  > General SMART Values:
                  > Offline data collection status: (0x82) Offline data collection activity
                  > was completed without error.
                  > Auto Offline Data Collection: Enabled.
                  > Self-test execution status: ( 0) The previous self-test routine
                  > completed without error or no self-test has ever
                  > been run.
                  > Total time to complete Offline
                  > data collection: ( 363) seconds.
                  > Offline data collection
                  > capabilities: (0x5b) SMART execute Offline immediate.
                  > Auto Offline data collection on/off support.
                  > Suspend Offline collection upon new
                  > command.
                  > Offline surface scan supported.
                  > Self-test supported.
                  > No Conveyance Self-test supported.
                  > Selective Self-test supported.
                  > SMART capabilities: (0x0003) Saves SMART data before entering
                  > power-saving mode.
                  > Supports SMART auto save timer.
                  > Error logging capability: (0x01) Error logging supported.
                  > No General Purpose Logging support.
                  > Short self-test routine
                  > recommended polling time: ( 2) minutes.
                  > Extended self-test routine
                  > recommended polling time: ( 107) minutes.
                  >
                  > SMART Attributes Data Structure revision number: 16
                  > Vendor Specific SMART Attributes with Thresholds:
                  > ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
                  > WHEN_FAILED RAW_VALUE 3 Spin_Up_Time 0x0027 175 175 063
                  > Pre-fail Always - 18693 4 Start_Stop_Count 0x0032 253
                  > 253 000 Old_age Always - 58 5 Reallocated_Sector_Ct
                  > 0x0033 253 253 063 Pre-fail Always - 0 6
                  > Read_Channel_Margin 0x0001 253 253 100 Pre-fail Offline
                  > - 0 7 Seek_Error_Rate 0x000a 253 252 000 Old_age
                  > Always - 0 8 Seek_Time_Performance 0x0027 251 246 187
                  > Pre-fail Always - 47887 9 Power_On_Minutes 0x0032
                  > 237 237 000 Old_age Always - 206h+53m 10
                  > Spin_Retry_Count 0x002b 253 252 157 Pre-fail Always
                  > - 0 11 Calibration_Retry_Count 0x002b 253 252 223 Pre-fail
                  > Always - 0 12 Power_Cycle_Count 0x0032 253 253 000
                  > Old_age Always - 81 192 Power-Off_Retract_Count 0x0032
                  > 253 253 000 Old_age Always - 0 193 Load_Cycle_Count
                  > 0x0032 253 253 000 Old_age Always - 0 194
                  > Temperature_Celsius 0x0032 253 253 000 Old_age Always
                  > - 40 195 Hardware_ECC_Recovered 0x000a 253 252 000 Old_age
                  > Always - 5485 196 Reallocated_Event_Count 0x0008 253 253
                  > 000 Old_age Offline - 0 197 Current_Pending_Sector 0x0008
                  > 253 253 000 Old_age Offline - 0 198
                  > Offline_Uncorrectable 0x0008 253 253 000 Old_age Offline
                  > - 0 199 UDMA_CRC_Error_Count 0x0008 199 199 000 Old_age
                  > Offline - 0 200 Multi_Zone_Error_Rate 0x000a 253 252 000
                  > Old_age Always - 0 201 Soft_Read_Error_Rate 0x000a
                  > 253 252 000 Old_age Always - 3 202 TA_Increase_Count
                  > 0x000a 253 252 000 Old_age Always - 0 203
                  > Run_Out_Cancel 0x000b 253 252 180 Pre-fail Always
                  > - 1 204 Shock_Count_Write_Opern 0x000a 253 252 000 Old_age
                  > Always - 0 205 Shock_Rate_Write_Opern 0x000a 253 252 000
                  > Old_age Always - 0 207 Spin_High_Current 0x002a
                  > 253 252 000 Old_age Always - 0 208 Spin_Buzz
                  > 0x002a 253 252 000 Old_age Always - 0 209
                  > Offline_Seek_Performnce 0x0024 196 189 000 Old_age Offline
                  > - 0 99 Unknown_Attribute 0x0004 253 253 000 Old_age
                  > Offline - 0 100 Unknown_Attribute 0x0004 253 253 000
                  > Old_age Offline - 0 101 Unknown_Attribute 0x0004
                  > 253 253 000 Old_age Offline - 0
                  >
                  > SMART Error Log Version: 1
                  > No Errors Logged
                  >
                  > SMART Self-test log structure revision number 1
                  > Num Test_Description Status Remaining LifeTime(hours)
                  > LBA_of_first_error # 1 Short offline Completed without error
                  > 00% 5309 - # 2 Extended offline Completed without error
                  > 00% 5309 - # 3 Short offline Completed without error
                  > 00% 5287 - # 4 Short offline Completed without
                  > error 00% 5265 - # 5 Short offline Completed
                  > without error 00% 5249 - # 6 Short offline
                  > Completed without error 00% 5220 - # 7 Short offline
                  > Completed without error 00% 5198 - # 8 Short offline
                  > Completed without error 00% 5175 - # 9 Short
                  > offline Completed without error 00% 5153 - #10
                  > Extended offline Completed without error 00% 5153 -
                  > #11 Short offline Completed without error 00% 5131
                  > - #12 Short offline Completed without error 00% 5108
                  > - #13 Short offline Completed without error 00% 5086
                  > - #14 Short offline Completed without error 00%
                  > 5063 - #15 Short offline Completed without error 00%
                  > 5041 - #16 Short offline Completed without error
                  > 00% 5019 - #17 Short offline Completed without error
                  > 00% 4996 - #18 Extended offline Completed without error
                  > 00% 4996 - #19 Short offline Completed without
                  > error 00% 4974 - #20 Short offline Completed
                  > without error 00% 4952 - #21 Short offline
                  > Completed without error 00% 4929 -
                  >
                  > SMART Selective self-test log data structure revision number 1
                  > SPAN MIN_LBA MAX_LBA CURRENT_TEST_STATUS
                  > 1 0 0 Not_testing
                  > 2 0 0 Not_testing
                  > 3 0 0 Not_testing
                  > 4 0 0 Not_testing
                  > 5 0 0 Not_testing
                  > Selective self-test flags (0x0):
                  > After scanning selected spans, do NOT read-scan remainder of disk.
                  > If Selective self-test is pending on power-up, resume after 0 minute delay.
                  >
                  > Disk - /dev/sdi
                  > smartctl version 5.32 Copyright (C) 2002-4 Bruce Allen
                  > Home page is http://smartmontools.sourceforge.net/
                  >
                  > === START OF INFORMATION SECTION ===
                  > Device Model: Maxtor 7Y250M0
                  > Serial Number: Y625W7ZE
                  > Firmware Version: YAR51EW0
                  > Device is: In smartctl database [for details use: -P show]
                  > ATA Version is: 7
                  > ATA Standard is: ATA/ATAPI-7 T13 1532D revision 0
                  > Local Time is: Sun Jan 2 09:44:45 2005 GST
                  > SMART support is: Availa<br/><br/>(Message over 64 KB, truncated)
                • Brad Campbell
                  ... I have no problem with that recomendation and qualification, but you did not state that. You simply mentioned Xeon. I have not had bad luck with VIA
                  Message 8 of 29 , Jan 1, 2005
                    Manu Abraham wrote:
                    > On Sun January 2 2005 10:04 am, Brad Campbell wrote:
                    >
                    >>Manu Abraham wrote:
                    >>
                    >>>BTW that is a 64 bit card, i think. Could not see the connectors too
                    >>>clearly.. You require a Xeon processor based machine to use that card in
                    >>>that case..
                    >>
                    >>Hey! Lets not get vendor specific here. There are plenty of Opteron boards
                    >>that will fit the bill also! :p)
                    >>
                    >
                    > The reason i got vendor specific was only one... If you go for an Opteron
                    > processor, stick to AMD PCI implementations rather than VIA
                    > implementations ..

                    I have no problem with that recomendation and qualification, but you did not state that. You simply
                    mentioned Xeon. I have not had bad luck with VIA chipsets, just suffered somewhat from lousy PCI
                    data transfer rates. There certainly are quirks with the PCI implementation but obtaining decent
                    hardware in Dubai seems to be an unattainable task. Every computer store sells the cheapest crap
                    they can get their hands on. I'm sick of walking into a store with a detailed shopping list and
                    having the know nothing behind the counter tell me that this motherboard over here is just as good
                    because they are unwilling to try and obtain the one I actually want. I know what I want, I know why
                    I want it and I now know why 50% of my hardware purchases are over the net from the USA. If it's not
                    on the shelf in front of them it's too hard. Don't worry, because the next pillock that walks in
                    will take the crap off their hands.

                    Hmmm sorry about the rant. I have just been through the process of trying to buy more hardware in
                    Dubai. Ended up ordering it from the US again. Far less frustration, and even after Fedex and 10%
                    Customs tarrif it's mostly cheaper!


                    >>Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5
                    >>(which just rocks by the way) but a rebuild takes me nearly 9 hours. If I
                    >>was on a pair of 8 port cards on PCI-X I'm sure it would be much quicker.
                    >>
                    >
                    > What if something bad again happened during the rebuild ? It is really
                    > dangerous to have very long periods of rebuilds .. For the same reason only
                    > Adaptec and other major manufacturers have battery backed up Hardware RAID
                    > cards ..

                    If you lose another drive during the rebuild then you are SOL I'm afraid. This is the reason my new
                    15 drive box will run Raid-6. At least it gives me a little more peace of mind.

                    Can someone give me a benchmark for a 2TB Drive Hardware Raid-5 rebuild? SCSI or ATA? Most of the
                    hardware Raid cards I have played with still have internal throughput bottlenecks and won't max out
                    all the drives during a rebuild. Having said that they are bound to do better than a 7MB/s rebuild
                    speed (which is what I get). Like I said, my solution is not optimal, but it's certainly cheap.

                    Can you put together a NAS SCSI 3TB Raid-6 box for 15k AED? I don't think so.

                    Oh, and check out the smart log on sdj. It has developed a few bad sectors and required a data
                    refresh to prompt re-allocation. Yes it took 9 hours, but I still had a functioning system. The only
                    issue that can damage a rebuild is another drive failure. Power failure, lockup, reset or other
                    nasties just cause the rebuild to start from scratch next time the array is activated. Much better
                    than no redundancy at all. Now all I need is hotplug on SATA so I can swap drives live.

                    mdadm --fail /dev/md0 /dev/sdj1
                    mdadm --remove /dev/md0 /dev/sdj1
                    mdadm --add /dev/md0 /dev/sdj1

                    >>Having said that, my read/write speeds are acceptable. Certainly not
                    >>lightning quick, but given I use it as a server for media streaming and
                    >>backups over Gig-E it does the job quite well. Price up 2TB of SCSI
                    >>sometime and check the tradeoff.
                    >
                    >
                    > Parallel SCSI is dead, the new implementations of SCSI are serial in nature,
                    > like iSCSI and so on.. Only for backward compatibility reasons parallel SCSI
                    > do exist for the time being ..

                    My comment remains.. Price up 2TB of SCSI (of any variety) and check the tradeoff.
                    And I (and many SCSI vendors) take some issue with the comment that Parallel SCSI is dead. It's
                    still pretty alive and kicking from where I stand. iSCSI is a completely different beast.
                    ATAoE looks interesting however.

                    Brad
                  • Manu Abraham
                    ... If you happen to have any media processing hardware, you are in bad luck with VIA, noted are also the Voodoo cards, permedia cards etc... ... Let s forget
                    Message 9 of 29 , Jan 2, 2005
                      On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                      > Manu Abraham wrote:
                      > > On Sun January 2 2005 10:04 am, Brad Campbell wrote:
                      > >>Manu Abraham wrote:
                      > >>>BTW that is a 64 bit card, i think. Could not see the connectors too
                      > >>>clearly.. You require a Xeon processor based machine to use that card in
                      > >>>that case..
                      > >>
                      > >>Hey! Lets not get vendor specific here. There are plenty of Opteron
                      > >> boards that will fit the bill also! :p)
                      > >
                      > > The reason i got vendor specific was only one... If you go for an Opteron
                      > > processor, stick to AMD PCI implementations rather than VIA
                      > > implementations ..
                      >
                      > I have no problem with that recomendation and qualification, but you did
                      > not state that. You simply mentioned Xeon. I have not had bad luck with VIA
                      If you happen to have any media processing hardware, you are in bad luck with
                      VIA, noted are also the Voodoo cards, permedia cards etc...

                      > chipsets, just suffered somewhat from lousy PCI data transfer rates. There
                      Let's forget the bad transfer .. One guy in UK could not find his DVB
                      hardware's frontend because he was using a VIA motherboard, some soya sauce
                      motherboard or something like that. oh, you have tomato sauce motherboards
                      also. Companies like these only have spoiled the AMD processor's market..

                      If you buy some motherboard at some point and later on find that, what you
                      want to does not work well with that board, then it is really sad. :-(


                      Eventhough a bit expensive if you are looking at the Opteron, You can pick up
                      AMD based boards from Tyan from Vinod Nair (MediaSys) in the Media City.
                      But again price-wise they are not cheap.. but very good implementations.. Very
                      good Linux support also.. The manufacturer itself provides drivers as
                      compared to the case as that of Intel's .. :-)

                      > certainly are quirks with the PCI implementation but obtaining decent
                      > hardware in Dubai seems to be an unattainable task. Every computer store
                      > sells the cheapest crap they can get their hands on. I'm sick of walking
                      > into a store with a detailed shopping list and having the know nothing
                      > behind the counter tell me that this motherboard over here is just as good
                      > because they are unwilling to try and obtain the one I actually want. I
                      > know what I want, I know why I want it and I now know why 50% of my
                      > hardware purchases are over the net from the USA. If it's not on the shelf
                      > in front of them it's too hard. Don't worry, because the next pillock that
                      > walks in will take the crap off their hands.
                      True, very true, This aspect i agree with you more than 100%, I don't know how
                      you define it in mathematics.

                      The reason is majority of the buyers, they just buy the things, just for
                      buying it, not for the need. They don't think even that it would work for
                      them.

                      >
                      > Hmmm sorry about the rant. I have just been through the process of trying
                      > to buy more hardware in Dubai. Ended up ordering it from the US again. Far
                      > less frustration, and even after Fedex and 10% Customs tarrif it's mostly
                      > cheaper!
                      Buy from UK/Germany/France, i mean Europe, it is a bit faster.. I do lot of
                      purchases from Germany. It is a bit faster than from US. US guys are a bit
                      lazy more than others..

                      >
                      > >>Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5
                      > >>(which just rocks by the way) but a rebuild takes me nearly 9 hours. If I
                      > >>was on a pair of 8 port cards on PCI-X I'm sure it would be much quicker.
                      > >
                      > > What if something bad again happened during the rebuild ? It is really
                      > > dangerous to have very long periods of rebuilds .. For the same reason
                      > > only Adaptec and other major manufacturers have battery backed up
                      > > Hardware RAID cards ..
                      >
                      > If you lose another drive during the rebuild then you are SOL I'm afraid.
                      > This is the reason my new 15 drive box will run Raid-6. At least it gives
                      > me a little more peace of mind.

                      rather than a disk failure, suppose you got a couple of bad superblocks where
                      the parity data resides. What do you do ? I mean on a restore .. Since during
                      a disk failure it is quite possible that drives from a batch are considered
                      quite bad..

                      Another point is that drives after shipping (i mean the total volume) that
                      approximately 10 ~ 13% is bad !, irrespective of manufacturer.
                      Although as years progressed, these figures have fluctuated due to lot of
                      reasons.

                      >
                      > Can someone give me a benchmark for a 2TB Drive Hardware Raid-5 rebuild?
                      > SCSI or ATA? Most of the hardware Raid cards I have played with still have
                      > internal throughput bottlenecks and won't max out all the drives during a
                      > rebuild. Having said that they are bound to do better than a 7MB/s rebuild
                      > speed (which is what I get). Like I said, my solution is not optimal, but
                      > it's certainly cheap.
                      >
                      > Can you put together a NAS SCSI 3TB Raid-6 box for 15k AED? I don't think
                      > so.
                      >
                      > Oh, and check out the smart log on sdj. It has developed a few bad sectors
                      > and required a data refresh to prompt re-allocation. Yes it took 9 hours,
                      > but I still had a functioning system. The only issue that can damage a
                      > rebuild is another drive failure. Power failure, lockup, reset or other
                      > nasties just cause the rebuild to start from scratch next time the array is
                      > activated. Much better than no redundancy at all. Now all I need is hotplug
                      > on SATA so I can swap drives live.
                      For Hot swap, you require drive cages, actually what these cages do is they
                      have a microcontroller to shutdown power to the drive before removal..

                      >
                      > mdadm --fail /dev/md0 /dev/sdj1
                      > mdadm --remove /dev/md0 /dev/sdj1
                      > mdadm --add /dev/md0 /dev/sdj1
                      >
                      > >>Having said that, my read/write speeds are acceptable. Certainly not
                      > >>lightning quick, but given I use it as a server for media streaming and
                      > >>backups over Gig-E it does the job quite well. Price up 2TB of SCSI
                      > >>sometime and check the tradeoff.
                      > >
                      > > Parallel SCSI is dead, the new implementations of SCSI are serial in
                      > > nature, like iSCSI and so on.. Only for backward compatibility reasons
                      > > parallel SCSI do exist for the time being ..
                      >
                      > My comment remains.. Price up 2TB of SCSI (of any variety) and check the
                      > tradeoff. And I (and many SCSI vendors) take some issue with the comment
                      > that Parallel SCSI is dead. It's still pretty alive and kicking from where
                      No i would say that the only reason why parallel SCSI is used for backward
                      compatibilty and _not_ meant for new installations, considering statements
                      from IBM, SGI and so on..

                      There will be stuff on the production line, but that is meant only for
                      backward compatibilty and so on.. You will not see anything new in parallel
                      SCSI. People call it deprecated as termed by Alan Cox.

                      You can buy new machines with parallel SCSI. I am not debating that..

                      > I stand. iSCSI is a completely different beast. ATAoE looks interesting
                      > however.
                      AoE as i said some time back, is promising, quite cheap also + hardware RAID
                      features also. Consider the box from coraid...

                      It is not just specifications as somebody else pointed out on the list .. It
                      _is_ working hardware. people make hardware only after specifications have
                      matured.. Otherwise those are called prototypes.. You don't get to buy
                      prototypes commercially, unless you are having a takeover of that product
                      design from that company.

                      Manu
                      >
                      > Brad
                      >
                      >
                      >
                      > Yahoo! Groups Links
                      >
                      >
                      >
                    • Manu Abraham
                      ... Try some good quality cards, Adaptec, Mylex (subsidiary of IBM), Intel etc.. I have worked with these cards, they have given me very good results, all
                      Message 10 of 29 , Jan 2, 2005
                        On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                        > Manu Abraham wrote:
                        > > On Sun January 2 2005 10:04 am, Brad Campbell wrote:
                        > >>Manu Abraham wrote:
                        > >>>BTW that is a 64 bit card, i think. Could not see the connectors too
                        > >>>clearly.. You require a Xeon processor based machine to use that card in
                        > >>>that case..
                        > >>
                        > >>Hey! Lets not get vendor specific here. There are plenty of Opteron
                        > >> boards that will fit the bill also! :p)
                        > >
                        > > The reason i got vendor specific was only one... If you go for an Opteron
                        > > processor, stick to AMD PCI implementations rather than VIA
                        > > implementations ..
                        >
                        > I have no problem with that recomendation and qualification, but you did
                        > not state that. You simply mentioned Xeon. I have not had bad luck with VIA
                        > chipsets, just suffered somewhat from lousy PCI data transfer rates. There
                        > certainly are quirks with the PCI implementation but obtaining decent
                        > hardware in Dubai seems to be an unattainable task. Every computer store
                        > sells the cheapest crap they can get their hands on. I'm sick of walking
                        > into a store with a detailed shopping list and having the know nothing
                        > behind the counter tell me that this motherboard over here is just as good
                        > because they are unwilling to try and obtain the one I actually want. I
                        > know what I want, I know why I want it and I now know why 50% of my
                        > hardware purchases are over the net from the USA. If it's not on the shelf
                        > in front of them it's too hard. Don't worry, because the next pillock that
                        > walks in will take the crap off their hands.
                        >
                        > Hmmm sorry about the rant. I have just been through the process of trying
                        > to buy more hardware in Dubai. Ended up ordering it from the US again. Far
                        > less frustration, and even after Fedex and 10% Customs tarrif it's mostly
                        > cheaper!
                        >
                        > >>Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5
                        > >>(which just rocks by the way) but a rebuild takes me nearly 9 hours. If I
                        > >>was on a pair of 8 port cards on PCI-X I'm sure it would be much quicker.
                        > >
                        > > What if something bad again happened during the rebuild ? It is really
                        > > dangerous to have very long periods of rebuilds .. For the same reason
                        > > only Adaptec and other major manufacturers have battery backed up
                        > > Hardware RAID cards ..
                        >
                        > If you lose another drive during the rebuild then you are SOL I'm afraid.
                        > This is the reason my new 15 drive box will run Raid-6. At least it gives
                        > me a little more peace of mind.
                        >
                        > Can someone give me a benchmark for a 2TB Drive Hardware Raid-5 rebuild?
                        > SCSI or ATA? Most of the hardware Raid cards I have played with still have
                        > internal throughput bottlenecks and won't max out all the drives during a
                        > rebuild. Having said that they are bound to do better than a 7MB/s rebuild
                        > speed (which is what I get). Like I said, my solution is not optimal, but
                        > it's certainly cheap.
                        Try some good quality cards, Adaptec, Mylex (subsidiary of IBM), Intel etc..
                        I have worked with these cards, they have given me very good results, all
                        these cards had a 64 bit processor and a minimum of 64MB of RAM, even the
                        oldest card that i came through..

                        Even though when somebody asks me about SCSI/SCSI-RAID cards, Adaptec is the
                        first one to jump out of my mouth ..

                        There is a lot of mathematics involved in how and when exactly, the redundant
                        data is written, so the firmware plays a lot of role in optimal throughputs..

                        Anyway you get only what you pay for..

                        Manu
                      • Brad Campbell
                        ... I have drive cages, but no Linux drivers yet relaibly support hot removal/addition of drives. This is something that is being worked on, but as yet has not
                        Message 11 of 29 , Jan 2, 2005
                          Manu Abraham wrote:
                          > On Sun January 2 2005 11:35 am, Brad Campbell wrote:

                          >>Oh, and check out the smart log on sdj. It has developed a few bad sectors
                          >>and required a data refresh to prompt re-allocation. Yes it took 9 hours,
                          >>but I still had a functioning system. The only issue that can damage a
                          >>rebuild is another drive failure. Power failure, lockup, reset or other
                          >>nasties just cause the rebuild to start from scratch next time the array is
                          >>activated. Much better than no redundancy at all. Now all I need is hotplug
                          >>on SATA so I can swap drives live.
                          >
                          > For Hot swap, you require drive cages, actually what these cages do is they
                          > have a microcontroller to shutdown power to the drive before removal..

                          I have drive cages, but no Linux drivers yet relaibly support hot removal/addition of drives. This
                          is something that is being worked on, but as yet has not matured enough even to make it to the
                          libata-dev tree. You can unofficially yank a disk and replace it with an "identical" unit, but if
                          you get a case like I have where I have 2 WD2000JB drives that are supposed to be identical, but due
                          to varying revisions actually have a different number of blocks, then nasty things might happen.
                          There is as yet no way to instruct the SATA layer to re-scan a disk.

                          SATA drives don't need any support other than a well built connector. The drives don't need to be
                          powered down in any particular sequence.

                          If you find you have bad blocks on other drives during a restore then you have problems. If you
                          check my smart logs you will note that all 14 drives in the machine undergo a short test every
                          morning and a long test once a week. This picks up any pending bad sectors to be rectified hopefully
                          before a catastrophic drive failure.
                          Having said that, if you do lose a stripe you are no worse off that developing bad sectors on a
                          non-redundant machine. The remainder of the data is going to be ok, just that one stripe is going to
                          be dud. You can even stop the array, dd the drive to a known working drive and do a forced array
                          start. (which you can't do with a hardware controller card). So Soft raid is a little more flexible
                          there.

                          Have a read of the linux-raid list. I have not seen anyone yet that says an ata or sata hardware
                          raid gets better performance than a well set up software raid across the board. There are selective
                          performance criteria where one will be better than the other however. I won't even mention SCSI as
                          it's just out of the ballpark price wise.
                          Also, don't forget 5 years down the track when your hardware raid controller is discontinued and
                          unsupported and it dies.

                          If you are after SATA drive cages, Supermicro make some excellent ones (and have a local rep who
                          usually has stock in Dubai who is actually helpful! I think they stock 3Ware controllers also). I
                          have 2x5 bay units in one machine and I have just ordered another 3x5 bay units for my NAS box.

                          Brad
                        • Manu Abraham
                          ... Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA hotswap.. I never tried SATA hotswap.. Hotswap i have only on my HA servers. ... I meant
                          Message 12 of 29 , Jan 2, 2005
                            On Sun January 2 2005 2:33 pm, Brad Campbell wrote:
                            > Manu Abraham wrote:
                            > > On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                            > >>Oh, and check out the smart log on sdj. It has developed a few bad
                            > >> sectors and required a data refresh to prompt re-allocation. Yes it took
                            > >> 9 hours, but I still had a functioning system. The only issue that can
                            > >> damage a rebuild is another drive failure. Power failure, lockup, reset
                            > >> or other nasties just cause the rebuild to start from scratch next time
                            > >> the array is activated. Much better than no redundancy at all. Now all I
                            > >> need is hotplug on SATA so I can swap drives live.
                            > >
                            > > For Hot swap, you require drive cages, actually what these cages do is
                            > > they have a microcontroller to shutdown power to the drive before
                            > > removal..
                            >
                            Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA hotswap.. I
                            never tried SATA hotswap.. Hotswap i have only on my HA servers.

                            > I have drive cages, but no Linux drivers yet relaibly support hot
                            > removal/addition of drives. This is something that is being worked on, but
                            > as yet has not matured enough even to make it to the libata-dev tree. You
                            > can unofficially yank a disk and replace it with an "identical" unit, but
                            > if you get a case like I have where I have 2 WD2000JB drives that are
                            > supposed to be identical, but due to varying revisions actually have a
                            > different number of blocks, then nasty things might happen. There is as yet
                            > no way to instruct the SATA layer to re-scan a disk.
                            >
                            > SATA drives don't need any support other than a well built connector. The
                            > drives don't need to be powered down in any particular sequence.
                            >
                            > If you find you have bad blocks on other drives during a restore then you
                            > have problems. If you check my smart logs you will note that all 14 drives
                            > in the machine undergo a short test every morning and a long test once a
                            > week. This picks up any pending bad sectors to be rectified hopefully
                            > before a catastrophic drive failure.
                            > Having said that, if you do lose a stripe you are no worse off that
                            > developing bad sectors on a non-redundant machine. The remainder of the
                            > data is going to be ok, just that one stripe is going to be dud. You can
                            I meant when a drive fails.. and have a bad superblock due to some reason...
                            It need not be a bad sector alone.. Murphy get's lot of his friends
                            together..

                            That's why for large systems, XFS for linux is considered to be Linux on
                            steroids.

                            > even stop the array, dd the drive to a known working drive and do a forced
                            > array start. (which you can't do with a hardware controller card). So Soft
                            > raid is a little more flexible there.
                            >
                            > Have a read of the linux-raid list. I have not seen anyone yet that says an
                            > ata or sata hardware raid gets better performance than a well set up
                            For an idling machine that is perfectly alright, Do you know how software RAID
                            works ? Idle CPU cycles.. But what if the CPU is not idle ? Having some
                            crunching work .. You loose performance, that's what you pay the price for
                            Software RAID.

                            Another one is that when the controller is talking to multiple drives, your
                            PCI bus can easily get flooded, but when a dedicated controller is talking to
                            the disk, the PCI bus is relatively free, processor also..

                            I believe you understand what i really mean..

                            If you are talking about something like the visualization systems we had,
                            software RAID is an utter NO .. No .. for the same said reasons ..

                            > software raid across the board. There are selective performance criteria
                            > where one will be better than the other however. I won't even mention SCSI
                            > as it's just out of the ballpark price wise.
                            > Also, don't forget 5 years down the track when your hardware raid
                            > controller is discontinued and unsupported and it dies.
                            It is still there check out Adaptec DPT-I20 64 bit SCSI RAID, after 5 years
                            still alive.. I am currently using the IPS 64bit SCSI-RAID, but i would say
                            the Adaptec was better than the IPS.

                            >
                            > If you are after SATA drive cages, Supermicro make some excellent ones (and
                            > have a local rep who usually has stock in Dubai who is actually helpful! I
                            > think they stock 3Ware controllers also). I have 2x5 bay units in one
                            > machine and I have just ordered another 3x5 bay units for my NAS box.
                            >
                            I prefer HA servers to be branded.

                            Manu
                            > Brad
                            >
                            >
                            >
                            > Yahoo! Groups Links
                            >
                            >
                            >
                          • Manu Abraham
                            ... try 203.200.16.128 ~ 150 128 and 130 have higher TTL s Manu
                            Message 13 of 29 , Jan 2, 2005
                              On Sun January 2 2005 2:33 pm, Brad Campbell wrote:
                              > Manu Abraham wrote:
                              > > On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                              > >>Oh, and check out the smart log on sdj. It has developed a few bad
                              > >> sectors and required a data refresh to prompt re-allocation. Yes it took
                              > >> 9 hours, but I still had a functioning system. The only issue that can
                              > >> damage a rebuild is another drive failure. Power failure, lockup, reset
                              > >> or other nasties just cause the rebuild to start from scratch next time
                              > >> the array is activated. Much better than no redundancy at all. Now all I
                              > >> need is hotplug on SATA so I can swap drives live.
                              > >
                              > > For Hot swap, you require drive cages, actually what these cages do is
                              > > they have a microcontroller to shutdown power to the drive before
                              > > removal..
                              >
                              > I have drive cages, but no Linux drivers yet relaibly support hot
                              > removal/addition of drives. This is something that is being worked on, but
                              > as yet has not matured enough even to make it to the libata-dev tree. You
                              > can unofficially yank a disk and replace it with an "identical" unit, but
                              > if you get a case like I have where I have 2 WD2000JB drives that are
                              > supposed to be identical, but due to varying revisions actually have a
                              > different number of blocks, then nasty things might happen. There is as yet
                              > no way to instruct the SATA layer to re-scan a disk.
                              >
                              > SATA drives don't need any support other than a well built connector. The
                              > drives don't need to be powered down in any particular sequence.
                              >
                              > If you find you have bad blocks on other drives during a restore then you
                              > have problems. If you check my smart logs you will note that all 14 drives
                              > in the machine undergo a short test every morning and a long test once a
                              > week. This picks up any pending bad sectors to be rectified hopefully
                              > before a catastrophic drive failure.
                              > Having said that, if you do lose a stripe you are no worse off that
                              > developing bad sectors on a non-redundant machine. The remainder of the
                              > data is going to be ok, just that one stripe is going to be dud. You can
                              > even stop the array, dd the drive to a known working drive and do a forced
                              > array start. (which you can't do with a hardware controller card). So Soft
                              > raid is a little more flexible there.
                              >
                              > Have a read of the linux-raid list. I have not seen anyone yet that says an
                              > ata or sata hardware raid gets better performance than a well set up
                              > software raid across the board. There are selective performance criteria
                              > where one will be better than the other however. I won't even mention SCSI
                              > as it's just out of the ballpark price wise.
                              > Also, don't forget 5 years down the track when your hardware raid
                              > controller is discontinued and unsupported and it dies.

                              try 203.200.16.128 ~ 150

                              128 and 130 have higher TTL's

                              Manu

                              >
                              > If you are after SATA drive cages, Supermicro make some excellent ones (and
                              > have a local rep who usually has stock in Dubai who is actually helpful! I
                              > think they stock 3Ware controllers also). I have 2x5 bay units in one
                              > machine and I have just ordered another 3x5 bay units for my NAS box.
                              >
                              > Brad
                              >
                              >
                              >
                              > Yahoo! Groups Links
                              >
                              >
                              >
                            • Manu Abraham
                              ... Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA hotswap.. I never tried SATA hotswap.. Hotswap i have only on my HA servers. ... I meant
                              Message 14 of 29 , Jan 2, 2005
                                On Sun January 2 2005 2:33 pm, Brad Campbell wrote:
                                > Manu Abraham wrote:
                                > > On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                                > >>Oh, and check out the smart log on sdj. It has developed a few bad
                                > >> sectors and required a data refresh to prompt re-allocation. Yes it took
                                > >> 9 hours, but I still had a functioning system. The only issue that can
                                > >> damage a rebuild is another drive failure. Power failure, lockup, reset
                                > >> or other nasties just cause the rebuild to start from scratch next time
                                > >> the array is activated. Much better than no redundancy at all. Now all I
                                > >> need is hotplug on SATA so I can swap drives live.
                                > >
                                > > For Hot swap, you require drive cages, actually what these cages do is
                                > > they have a microcontroller to shutdown power to the drive before
                                > > removal..
                                >
                                Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA hotswap.. I
                                never tried SATA hotswap.. Hotswap i have only on my HA servers.

                                > I have drive cages, but no Linux drivers yet relaibly support hot
                                > removal/addition of drives. This is something that is being worked on, but
                                > as yet has not matured enough even to make it to the libata-dev tree. You
                                > can unofficially yank a disk and replace it with an "identical" unit, but
                                > if you get a case like I have where I have 2 WD2000JB drives that are
                                > supposed to be identical, but due to varying revisions actually have a
                                > different number of blocks, then nasty things might happen. There is as yet
                                > no way to instruct the SATA layer to re-scan a disk.
                                >
                                > SATA drives don't need any support other than a well built connector. The
                                > drives don't need to be powered down in any particular sequence.
                                >
                                > If you find you have bad blocks on other drives during a restore then you
                                > have problems. If you check my smart logs you will note that all 14 drives
                                > in the machine undergo a short test every morning and a long test once a
                                > week. This picks up any pending bad sectors to be rectified hopefully
                                > before a catastrophic drive failure.
                                > Having said that, if you do lose a stripe you are no worse off that
                                > developing bad sectors on a non-redundant machine. The remainder of the
                                > data is going to be ok, just that one stripe is going to be dud. You can
                                I meant when a drive fails.. and have a bad superblock due to some reason...
                                It need not be a bad sector alone.. Murphy get's lot of his friends
                                together..

                                That's why for large systems, XFS for linux is considered to be Linux on
                                steroids.

                                > even stop the array, dd the drive to a known working drive and do a forced
                                > array start. (which you can't do with a hardware controller card). So Soft
                                > raid is a little more flexible there.
                                >
                                > Have a read of the linux-raid list. I have not seen anyone yet that says an
                                > ata or sata hardware raid gets better performance than a well set up
                                For an idling machine that is perfectly alright, Do you know how software RAID
                                works ? Idle CPU cycles.. But what if the CPU is not idle ? Having some
                                crunching work .. You loose performance, that's what you pay the price for
                                Software RAID.

                                Another one is that when the controller is talking to multiple drives, your
                                PCI bus can easily get flooded, but when a dedicated controller is talking to
                                the disk, the PCI bus is relatively free, processor also..

                                I believe you understand what i really mean..

                                If you are talking about something like the visualization systems we had,
                                software RAID is an utter NO .. No .. for the same said reasons ..

                                > software raid across the board. There are selective performance criteria
                                > where one will be better than the other however. I won't even mention SCSI
                                > as it's just out of the ballpark price wise.
                                > Also, don't forget 5 years down the track when your hardware raid
                                > controller is discontinued and unsupported and it dies.
                                It is still there check out Adaptec DPT-I20 64 bit SCSI RAID, after 5 years
                                still alive.. I am currently using the IPS 64bit SCSI-RAID, but i would say
                                the Adaptec was better than the IPS.

                                >
                                > If you are after SATA drive cages, Supermicro make some excellent ones (and
                                > have a local rep who usually has stock in Dubai who is actually helpful! I
                                > think they stock 3Ware controllers also). I have 2x5 bay units in one
                                > machine and I have just ordered another 3x5 bay units for my NAS box.
                                >
                                I prefer HA servers to be branded.

                                Manu
                                > Brad
                                >
                                >
                                >
                                > Yahoo! Groups Links
                                >
                                >
                                >
                              • Manu Abraham
                                ... If you happen to have any media processing hardware, you are in bad luck with VIA, noted are also the Voodoo cards, permedia cards etc... ... Let s forget
                                Message 15 of 29 , Jan 2, 2005
                                  On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                                  > Manu Abraham wrote:
                                  > > On Sun January 2 2005 10:04 am, Brad Campbell wrote:
                                  > >>Manu Abraham wrote:
                                  > >>>BTW that is a 64 bit card, i think. Could not see the connectors too
                                  > >>>clearly.. You require a Xeon processor based machine to use that card in
                                  > >>>that case..
                                  > >>
                                  > >>Hey! Lets not get vendor specific here. There are plenty of Opteron
                                  > >> boards that will fit the bill also! :p)
                                  > >
                                  > > The reason i got vendor specific was only one... If you go for an Opteron
                                  > > processor, stick to AMD PCI implementations rather than VIA
                                  > > implementations ..
                                  >
                                  > I have no problem with that recomendation and qualification, but you did
                                  > not state that. You simply mentioned Xeon. I have not had bad luck with VIA
                                  If you happen to have any media processing hardware, you are in bad luck with
                                  VIA, noted are also the Voodoo cards, permedia cards etc...

                                  > chipsets, just suffered somewhat from lousy PCI data transfer rates. There
                                  Let's forget the bad transfer .. One guy in UK could not find his DVB
                                  hardware's frontend because he was using a VIA motherboard, some soya sauce
                                  motherboard or something like that. oh, you have tomato sauce motherboards
                                  also. Companies like these only have spoiled the AMD processor's market..

                                  If you buy some motherboard at some point and later on find that, what you
                                  want to does not work well with that board, then it is really sad. :-(


                                  Eventhough a bit expensive if you are looking at the Opteron, You can pick up
                                  AMD based boards from Tyan from Vinod Nair (MediaSys) in the Media City.
                                  But again price-wise they are not cheap.. but very good implementations.. Very
                                  good Linux support also.. The manufacturer itself provides drivers as
                                  compared to the case as that of Intel's .. :-)

                                  > certainly are quirks with the PCI implementation but obtaining decent
                                  > hardware in Dubai seems to be an unattainable task. Every computer store
                                  > sells the cheapest crap they can get their hands on. I'm sick of walking
                                  > into a store with a detailed shopping list and having the know nothing
                                  > behind the counter tell me that this motherboard over here is just as good
                                  > because they are unwilling to try and obtain the one I actually want. I
                                  > know what I want, I know why I want it and I now know why 50% of my
                                  > hardware purchases are over the net from the USA. If it's not on the shelf
                                  > in front of them it's too hard. Don't worry, because the next pillock that
                                  > walks in will take the crap off their hands.
                                  True, very true, This aspect i agree with you more than 100%, I don't know how
                                  you define it in mathematics.

                                  The reason is majority of the buyers, they just buy the things, just for
                                  buying it, not for the need. They don't think even that it would work for
                                  them.

                                  >
                                  > Hmmm sorry about the rant. I have just been through the process of trying
                                  > to buy more hardware in Dubai. Ended up ordering it from the US again. Far
                                  > less frustration, and even after Fedex and 10% Customs tarrif it's mostly
                                  > cheaper!
                                  Buy from UK/Germany/France, i mean Europe, it is a bit faster.. I do lot of
                                  purchases from Germany. It is a bit faster than from US. US guys are a bit
                                  lazy more than others..

                                  >
                                  > >>Agreed. I have 2TB on 3 PCI cards and 10 drives running software RAID-5
                                  > >>(which just rocks by the way) but a rebuild takes me nearly 9 hours. If I
                                  > >>was on a pair of 8 port cards on PCI-X I'm sure it would be much quicker.
                                  > >
                                  > > What if something bad again happened during the rebuild ? It is really
                                  > > dangerous to have very long periods of rebuilds .. For the same reason
                                  > > only Adaptec and other major manufacturers have battery backed up
                                  > > Hardware RAID cards ..
                                  >
                                  > If you lose another drive during the rebuild then you are SOL I'm afraid.
                                  > This is the reason my new 15 drive box will run Raid-6. At least it gives
                                  > me a little more peace of mind.

                                  rather than a disk failure, suppose you got a couple of bad superblocks where
                                  the parity data resides. What do you do ? I mean on a restore .. Since during
                                  a disk failure it is quite possible that drives from a batch are considered
                                  quite bad..

                                  Another point is that drives after shipping (i mean the total volume) that
                                  approximately 10 ~ 13% is bad !, irrespective of manufacturer.
                                  Although as years progressed, these figures have fluctuated due to lot of
                                  reasons.

                                  >
                                  > Can someone give me a benchmark for a 2TB Drive Hardware Raid-5 rebuild?
                                  > SCSI or ATA? Most of the hardware Raid cards I have played with still have
                                  > internal throughput bottlenecks and won't max out all the drives during a
                                  > rebuild. Having said that they are bound to do better than a 7MB/s rebuild
                                  > speed (which is what I get). Like I said, my solution is not optimal, but
                                  > it's certainly cheap.
                                  >
                                  > Can you put together a NAS SCSI 3TB Raid-6 box for 15k AED? I don't think
                                  > so.
                                  >
                                  > Oh, and check out the smart log on sdj. It has developed a few bad sectors
                                  > and required a data refresh to prompt re-allocation. Yes it took 9 hours,
                                  > but I still had a functioning system. The only issue that can damage a
                                  > rebuild is another drive failure. Power failure, lockup, reset or other
                                  > nasties just cause the rebuild to start from scratch next time the array is
                                  > activated. Much better than no redundancy at all. Now all I need is hotplug
                                  > on SATA so I can swap drives live.
                                  For Hot swap, you require drive cages, actually what these cages do is they
                                  have a microcontroller to shutdown power to the drive before removal..

                                  >
                                  > mdadm --fail /dev/md0 /dev/sdj1
                                  > mdadm --remove /dev/md0 /dev/sdj1
                                  > mdadm --add /dev/md0 /dev/sdj1
                                  >
                                  > >>Having said that, my read/write speeds are acceptable. Certainly not
                                  > >>lightning quick, but given I use it as a server for media streaming and
                                  > >>backups over Gig-E it does the job quite well. Price up 2TB of SCSI
                                  > >>sometime and check the tradeoff.
                                  > >
                                  > > Parallel SCSI is dead, the new implementations of SCSI are serial in
                                  > > nature, like iSCSI and so on.. Only for backward compatibility reasons
                                  > > parallel SCSI do exist for the time being ..
                                  >
                                  > My comment remains.. Price up 2TB of SCSI (of any variety) and check the
                                  > tradeoff. And I (and many SCSI vendors) take some issue with the comment
                                  > that Parallel SCSI is dead. It's still pretty alive and kicking from where
                                  No i would say that the only reason why parallel SCSI is used for backward
                                  compatibilty and _not_ meant for new installations, considering statements
                                  from IBM, SGI and so on..

                                  There will be stuff on the production line, but that is meant only for
                                  backward compatibilty and so on.. You will not see anything new in parallel
                                  SCSI. People call it deprecated as termed by Alan Cox.

                                  You can buy new machines with parallel SCSI. I am not debating that..

                                  > I stand. iSCSI is a completely different beast. ATAoE looks interesting
                                  > however.
                                  AoE as i said some time back, is promising, quite cheap also + hardware RAID
                                  features also. Consider the box from coraid...

                                  It is not just specifications as somebody else pointed out on the list .. It
                                  _is_ working hardware. people make hardware only after specifications have
                                  matured.. Otherwise those are called prototypes.. You don't get to buy
                                  prototypes commercially, unless you are having a takeover of that product
                                  design from that company.

                                  Manu
                                  >
                                  > Brad
                                  >
                                  >
                                  >
                                  > Yahoo! Groups Links
                                  >
                                  >
                                  >
                                • Brad Campbell
                                  ... Fair enough, SCSI requires some tricky termination issues when removing/adding devices to the bus live. Yes other trickery is required. ... Murphy is a
                                  Message 16 of 29 , Jan 2, 2005
                                    Manu Abraham wrote:
                                    >>
                                    > Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA hotswap.. I
                                    > never tried SATA hotswap.. Hotswap i have only on my HA servers.

                                    Fair enough, SCSI requires some tricky termination issues when removing/adding devices to the bus
                                    live. Yes other trickery is required.

                                    >>Having said that, if you do lose a stripe you are no worse off that
                                    >>developing bad sectors on a non-redundant machine. The remainder of the
                                    >>data is going to be ok, just that one stripe is going to be dud. You can
                                    >
                                    > I meant when a drive fails.. and have a bad superblock due to some reason...
                                    > It need not be a bad sector alone.. Murphy get's lot of his friends
                                    > together..

                                    Murphy is a bastard!

                                    > That's why for large systems, XFS for linux is considered to be Linux on
                                    > steroids.

                                    Perhaps. Heard too many stories about XFS data loss. I stick with Ext-3. Certainly not the best
                                    choice, but certainly well tested, reliable and not the worst.

                                    > For an idling machine that is perfectly alright, Do you know how software RAID
                                    > works ? Idle CPU cycles.. But what if the CPU is not idle ? Having some
                                    > crunching work .. You loose performance, that's what you pay the price for
                                    > Software RAID.

                                    That is why I build NAS boxes. The server does all the crunching, the NAS boxes just manage the
                                    RAID. It still works out cheaper and more reliable in the long run.
                                    On the flip side, you lose write performance, but provided the array is not degraded read
                                    performance is nearly as good as RAID-0 as no calculation takes place at all.

                                    > Another one is that when the controller is talking to multiple drives, your
                                    > PCI bus can easily get flooded, but when a dedicated controller is talking to
                                    > the disk, the PCI bus is relatively free, processor also..
                                    >
                                    > I believe you understand what i really mean..

                                    I do.

                                    > If you are talking about something like the visualization systems we had,
                                    > software RAID is an utter NO .. No .. for the same said reasons ..
                                    >

                                    True, but if you are talking about high performance systems then any form of raid-5 is a no-no.

                                    >>If you are after SATA drive cages, Supermicro make some excellent ones (and
                                    >>have a local rep who usually has stock in Dubai who is actually helpful! I
                                    >>think they stock 3Ware controllers also). I have 2x5 bay units in one
                                    >>machine and I have just ordered another 3x5 bay units for my NAS box.
                                    >>
                                    >
                                    > I prefer HA servers to be branded.
                                    >

                                    So do I, but sometimes budget dictates otherwise. I had 1000AED to build the office fileserver,
                                    which dictates cheap components and hand built/configured machines. It's still RAID-1 and has not so
                                    much as farted in the last 18 months. We keep good backups in any case, but still, you can't argue
                                    with 1000AED to run a 25 PC network with full mail/fileserver/printserver/nat/firewall/web/gateway
                                    services.

                                    I'm tired of this conversation. I do it my way, on a budget. You do it yours.
                                    You probably make money from your machines. Mine hold together a small company on a budget (the
                                    1000AED server) and power my Home entertainment system.

                                    Brad
                                  • Manu Abraham
                                    ... Intelligent Input/Output (I2O) is a hardware specification that describes a model for offloading I/O processing from the CPU. There is no doubt that this
                                    Message 17 of 29 , Jan 2, 2005
                                      On Sun January 2 2005 2:33 pm, Brad Campbell wrote:
                                      > Manu Abraham wrote:
                                      > > On Sun January 2 2005 11:35 am, Brad Campbell wrote:
                                      > >>Oh, and check out the smart log on sdj. It has developed a few bad
                                      > >> sectors and required a data refresh to prompt re-allocation. Yes it took
                                      > >> 9 hours, but I still had a functioning system. The only issue that can
                                      > >> damage a rebuild is another drive failure. Power failure, lockup, reset
                                      > >> or other nasties just cause the rebuild to start from scratch next time
                                      > >> the array is activated. Much better than no redundancy at all. Now all I
                                      > >> need is hotplug on SATA so I can swap drives live.
                                      > >
                                      > > For Hot swap, you require drive cages, actually what these cages do is
                                      > > they have a microcontroller to shutdown power to the drive before
                                      > > removal..
                                      >
                                      > I have drive cages, but no Linux drivers yet relaibly support hot
                                      > removal/addition of drives. This is something that is being worked on, but
                                      > as yet has not matured enough even to make it to the libata-dev tree. You
                                      > can unofficially yank a disk and replace it with an "identical" unit, but
                                      > if you get a case like I have where I have 2 WD2000JB drives that are
                                      > supposed to be identical, but due to varying revisions actually have a
                                      > different number of blocks, then nasty things might happen. There is as yet
                                      > no way to instruct the SATA layer to re-scan a disk.
                                      >
                                      > SATA drives don't need any support other than a well built connector. The
                                      > drives don't need to be powered down in any particular sequence.
                                      >
                                      > If you find you have bad blocks on other drives during a restore then you
                                      > have problems. If you check my smart logs you will note that all 14 drives
                                      > in the machine undergo a short test every morning and a long test once a
                                      > week. This picks up any pending bad sectors to be rectified hopefully
                                      > before a catastrophic drive failure.
                                      > Having said that, if you do lose a stripe you are no worse off that
                                      > developing bad sectors on a non-redundant machine. The remainder of the
                                      > data is going to be ok, just that one stripe is going to be dud. You can
                                      > even stop the array, dd the drive to a known working drive and do a forced
                                      > array start. (which you can't do with a hardware controller card). So Soft
                                      > raid is a little more flexible there.
                                      >
                                      > Have a read of the linux-raid list. I have not seen anyone yet that says an
                                      > ata or sata hardware raid gets better performance than a well set up
                                      > software raid across the board. There are selective performance criteria

                                      Intelligent Input/Output (I2O) is a hardware specification that describes a
                                      model for offloading I/O processing from the CPU. There is no doubt that this
                                      is leading to extremely high performance I/O. The model is after the style of
                                      what has been used in very large mainframes for years. It is not a
                                      replacement for the PCI architecture.

                                      http://www.uni-mainz.de/~neuffer/i2o/what_is_i20.html

                                      Manu
                                      > where one will be better than the other however. I won't even mention SCSI
                                      > as it's just out of the ballpark price wise.
                                      > Also, don't forget 5 years down the track when your hardware raid
                                      > controller is discontinued and unsupported and it dies.
                                      >
                                      > If you are after SATA drive cages, Supermicro make some excellent ones (and
                                      > have a local rep who usually has stock in Dubai who is actually helpful! I
                                      > think they stock 3Ware controllers also). I have 2x5 bay units in one
                                      > machine and I have just ordered another 3x5 bay units for my NAS box.
                                      >
                                      > Brad
                                      >
                                      >
                                      >
                                      > Yahoo! Groups Links
                                      >
                                      >
                                      >
                                    • Manu Abraham
                                      ... No, not for SCA SCSI.. I mean Hot swappable ones.. I can demonstrate it sometime for you.. ... I agree.. ... Brad, i used to work in an organization with
                                      Message 18 of 29 , Jan 2, 2005
                                        On Sun January 2 2005 7:14 pm, Brad Campbell wrote:
                                        > Manu Abraham wrote:
                                        > > Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA
                                        > > hotswap.. I never tried SATA hotswap.. Hotswap i have only on my HA
                                        > > servers.
                                        >
                                        > Fair enough, SCSI requires some tricky termination issues when
                                        > removing/adding devices to the bus live. Yes other trickery is required.
                                        >
                                        No, not for SCA SCSI.. I mean Hot swappable ones.. I can demonstrate it
                                        sometime for you..

                                        > >>Having said that, if you do lose a stripe you are no worse off that
                                        > >>developing bad sectors on a non-redundant machine. The remainder of the
                                        > >>data is going to be ok, just that one stripe is going to be dud. You can
                                        > >
                                        > > I meant when a drive fails.. and have a bad superblock due to some
                                        > > reason... It need not be a bad sector alone.. Murphy get's lot of his
                                        > > friends together..
                                        >
                                        > Murphy is a bastard!
                                        I agree..
                                        >
                                        > > That's why for large systems, XFS for linux is considered to be Linux on
                                        > > steroids.
                                        >
                                        > Perhaps. Heard too many stories about XFS data loss. I stick with Ext-3.
                                        > Certainly not the best choice, but certainly well tested, reliable and not
                                        > the worst.
                                        >
                                        Brad, i used to work in an organization with around 60 SGI machines and having
                                        extreme volumes of data.. all CAE/CAM files are really huge. XFS was a
                                        solution.. i mean at that point of time..
                                        And was rock solid.. And that was 2.4.

                                        > > For an idling machine that is perfectly alright, Do you know how software
                                        > > RAID works ? Idle CPU cycles.. But what if the CPU is not idle ? Having
                                        > > some crunching work .. You loose performance, that's what you pay the
                                        > > price for Software RAID.
                                        >
                                        > That is why I build NAS boxes. The server does all the crunching, the NAS
                                        > boxes just manage the RAID. It still works out cheaper and more reliable in
                                        > the long run. On the flip side, you lose write performance, but provided
                                        > the array is not degraded read performance is nearly as good as RAID-0 as
                                        > no calculation takes place at all.
                                        Under the circumstances that i had even the network would be congested ...
                                        Running a PVM cluster with around 140 PIII nodes.
                                        >
                                        > > Another one is that when the controller is talking to multiple drives,
                                        > > your PCI bus can easily get flooded, but when a dedicated controller is
                                        > > talking to the disk, the PCI bus is relatively free, processor also..
                                        > >
                                        > > I believe you understand what i really mean..
                                        >
                                        > I do.
                                        >
                                        > > If you are talking about something like the visualization systems we had,
                                        > > software RAID is an utter NO .. No .. for the same said reasons ..
                                        >
                                        > True, but if you are talking about high performance systems then any form
                                        > of raid-5 is a no-no.
                                        That is why Hardware RAID is used .. All SGI machines what we had were all
                                        U320, while the Linux servers were running U160 DPT I2O..

                                        If you understand what I20 is, it gives you lot of performance boost.. In the
                                        sense that all processing is offloaded to the card's processor. This card is
                                        not very expensive.. It is almost the same price.. Only the SCSI drives are
                                        more expensive..

                                        >
                                        > >>If you are after SATA drive cages, Supermicro make some excellent ones
                                        > >> (and have a local rep who usually has stock in Dubai who is actually
                                        > >> helpful! I think they stock 3Ware controllers also). I have 2x5 bay
                                        > >> units in one machine and I have just ordered another 3x5 bay units for
                                        > >> my NAS box.
                                        > >
                                        > > I prefer HA servers to be branded.
                                        >
                                        > So do I, but sometimes budget dictates otherwise. I had 1000AED to build
                                        > the office fileserver, which dictates cheap components and hand
                                        True..

                                        > built/configured machines. It's still RAID-1 and has not so much as farted
                                        > in the last 18 months. We keep good backups in any case, but still, you
                                        > can't argue with 1000AED to run a 25 PC network with full
                                        > mail/fileserver/printserver/nat/firewall/web/gateway services.
                                        >
                                        > I'm tired of this conversation. I do it my way, on a budget. You do it
                                        Ok let's stop the conversation here..


                                        Manu
                                      • Rajeev
                                        Dear Manu and Brad, Thanks for your support and sharing your knowledge. It is really informative. Before you stop this conversation could you please give me a
                                        Message 19 of 29 , Jan 2, 2005
                                          Dear Manu and Brad,

                                          Thanks for your support and sharing your knowledge. It is really
                                          informative.

                                          Before you stop this conversation could you please give me a part list (with
                                          manufacturer name and model numbers if possible) of the NAS server you had
                                          build with 10 drives. I would like to build a similar one with 6 drives and
                                          increase it later as and when the data increases. Also I thought generally
                                          the hardware price are going down and there is no point in buying too much
                                          hardware for a future use at higher price!

                                          I was trying to get the price for the last 2 days, as you know most of the
                                          people whom I contacted where sales people and getting technical suggestion
                                          was difficult. Most of them talk in terms of MS point of view.

                                          If not too much can I have your contact numbers and convenient time to speak
                                          if you don't mind? .

                                          Regards,

                                          Rajeev
                                          O50-5247785


                                          -----Original Message-----
                                          From: Manu Abraham [mailto:manu@...]
                                          Sent: 02 January 2005 20:51
                                          To: dubailug@yahoogroups.com
                                          Subject: Re: [dubailug] Sata Drives


                                          On Sun January 2 2005 7:14 pm, Brad Campbell wrote:
                                          > Manu Abraham wrote:
                                          > > Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA
                                          > > hotswap.. I never tried SATA hotswap.. Hotswap i have only on my HA
                                          > > servers.
                                          >
                                          > Fair enough, SCSI requires some tricky termination issues when
                                          > removing/adding devices to the bus live. Yes other trickery is required.
                                          >
                                          No, not for SCA SCSI.. I mean Hot swappable ones.. I can demonstrate it
                                          sometime for you..

                                          > >>Having said that, if you do lose a stripe you are no worse off that
                                          > >>developing bad sectors on a non-redundant machine. The remainder of the
                                          > >>data is going to be ok, just that one stripe is going to be dud. You can
                                          > >
                                          > > I meant when a drive fails.. and have a bad superblock due to some
                                          > > reason... It need not be a bad sector alone.. Murphy get's lot of his
                                          > > friends together..
                                          >
                                          > Murphy is a bastard!
                                          I agree..
                                          >
                                          > > That's why for large systems, XFS for linux is considered to be Linux on
                                          > > steroids.
                                          >
                                          > Perhaps. Heard too many stories about XFS data loss. I stick with Ext-3.
                                          > Certainly not the best choice, but certainly well tested, reliable and not
                                          > the worst.
                                          >
                                          Brad, i used to work in an organization with around 60 SGI machines and
                                          having
                                          extreme volumes of data.. all CAE/CAM files are really huge. XFS was a
                                          solution.. i mean at that point of time..
                                          And was rock solid.. And that was 2.4.

                                          > > For an idling machine that is perfectly alright, Do you know how
                                          software
                                          > > RAID works ? Idle CPU cycles.. But what if the CPU is not idle ? Having
                                          > > some crunching work .. You loose performance, that's what you pay the
                                          > > price for Software RAID.
                                          >
                                          > That is why I build NAS boxes. The server does all the crunching, the NAS
                                          > boxes just manage the RAID. It still works out cheaper and more reliable
                                          in
                                          > the long run. On the flip side, you lose write performance, but provided
                                          > the array is not degraded read performance is nearly as good as RAID-0 as
                                          > no calculation takes place at all.
                                          Under the circumstances that i had even the network would be congested ...
                                          Running a PVM cluster with around 140 PIII nodes.
                                          >
                                          > > Another one is that when the controller is talking to multiple drives,
                                          > > your PCI bus can easily get flooded, but when a dedicated controller is
                                          > > talking to the disk, the PCI bus is relatively free, processor also..
                                          > >
                                          > > I believe you understand what i really mean..
                                          >
                                          > I do.
                                          >
                                          > > If you are talking about something like the visualization systems we
                                          had,
                                          > > software RAID is an utter NO .. No .. for the same said reasons ..
                                          >
                                          > True, but if you are talking about high performance systems then any form
                                          > of raid-5 is a no-no.
                                          That is why Hardware RAID is used .. All SGI machines what we had were all
                                          U320, while the Linux servers were running U160 DPT I2O..

                                          If you understand what I20 is, it gives you lot of performance boost.. In
                                          the
                                          sense that all processing is offloaded to the card's processor. This card is
                                          not very expensive.. It is almost the same price.. Only the SCSI drives are
                                          more expensive..

                                          >
                                          > >>If you are after SATA drive cages, Supermicro make some excellent ones
                                          > >> (and have a local rep who usually has stock in Dubai who is actually
                                          > >> helpful! I think they stock 3Ware controllers also). I have 2x5 bay
                                          > >> units in one machine and I have just ordered another 3x5 bay units for
                                          > >> my NAS box.
                                          > >
                                          > > I prefer HA servers to be branded.
                                          >
                                          > So do I, but sometimes budget dictates otherwise. I had 1000AED to build
                                          > the office fileserver, which dictates cheap components and hand
                                          True..

                                          > built/configured machines. It's still RAID-1 and has not so much as farted
                                          > in the last 18 months. We keep good backups in any case, but still, you
                                          > can't argue with 1000AED to run a 25 PC network with full
                                          > mail/fileserver/printserver/nat/firewall/web/gateway services.
                                          >
                                          > I'm tired of this conversation. I do it my way, on a budget. You do it
                                          Ok let's stop the conversation here..


                                          Manu



                                          Yahoo! Groups Links
                                        • John Joseph
                                          bparks64 Brian Parks brianp@gulfsolution.com Dubai christopherstuartsaul christopher.saul@sun.com dandantheitman Gentoo Dan (26/M) dandantheitman@gmail.com
                                          Message 20 of 29 , Jan 2, 2005
                                          • Brad Campbell
                                            ... I used an ASUS A7V600 Motherboard Athlon 2600+ Processor 3 x Promise SATA150TX4 4-Port PCI SATA controllers
                                            Message 21 of 29 , Jan 2, 2005
                                              Rajeev wrote:
                                              > Dear Manu and Brad,
                                              >
                                              > Thanks for your support and sharing your knowledge. It is really
                                              > informative.
                                              >
                                              > Before you stop this conversation could you please give me a part list (with
                                              > manufacturer name and model numbers if possible) of the NAS server you had
                                              > build with 10 drives. I would like to build a similar one with 6 drives and
                                              > increase it later as and when the data increases. Also I thought generally
                                              > the hardware price are going down and there is no point in buying too much
                                              > hardware for a future use at higher price!
                                              >

                                              I used an ASUS A7V600 Motherboard
                                              Athlon 2600+ Processor

                                              3 x Promise SATA150TX4 4-Port PCI SATA controllers
                                              http://www.promise.com/product/product_detail_eng.asp?productId=98&familyId=3

                                              512mb of DDR Ram

                                              12 x Maxtor 7Y250M0 Drives
                                              http://www.maxtor.com/portal/site/Maxtor/menuitem.ba88f6d7cf664718376049b291346068/?channelpath=/en_us/Products/ATA%20Hard%20Drives/Midline%20Applications/MaXLine%20Plus%20II&productview=Features

                                              2 x Supermicro CSE-M35T 5 Bay hotswap units
                                              http://www.supermicro.com/products/accessories/mobilerack/CSE-M35T-1.cfm

                                              The Server http://www.wasp.net.au/~brad/p1000250.jpg
                                              The PSU http://www.wasp.net.au/~brad/p1000256.jpg

                                              As Manu said. There are quirks with the PCI bridge on the ASUS that cause the odd slow down when
                                              reading/writing from multiple controllers at once. I don't have a problem with it, but I would
                                              recommend an AMD chipset board (I'm and anti Intel biggot). Or use an Intel board and processor (I'd
                                              rather attend a Kenny G concert!)

                                              Brad
                                            • Manu Abraham
                                              ... You can buy the SATA card that you had in mind, the Adaptec card, you can put it onto an ASUS Server/workstation class board, RAM and processor as per your
                                              Message 22 of 29 , Jan 2, 2005
                                                On Sun January 2 2005 10:24 pm, Rajeev wrote:
                                                > Dear Manu and Brad,
                                                >
                                                > Thanks for your support and sharing your knowledge. It is really
                                                > informative.
                                                >
                                                > Before you stop this conversation could you please give me a part list
                                                > (with manufacturer name and model numbers if possible) of the NAS server
                                                > you had build with 10 drives. I would like to build a similar one with 6
                                                > drives and increase it later as and when the data increases. Also I thought
                                                > generally the hardware price are going down and there is no point in buying
                                                > too much hardware for a future use at higher price!
                                                >
                                                > I was trying to get the price for the last 2 days, as you know most of the
                                                > people whom I contacted where sales people and getting technical suggestion
                                                > was difficult. Most of them talk in terms of MS point of view.
                                                >
                                                You can buy the SATA card that you had in mind, the Adaptec card, you can put
                                                it onto an ASUS Server/workstation class board, RAM and processor as per your
                                                requirement, if you are looking at NAS storage, hope you have sufficient N/w
                                                bandwidth, to handle the amount of data.

                                                If you are looking at a feature rich motherboard, you can go for the ASUS,
                                                NRL-LS533/NRL-L533.
                                                The advantage/features, would be you have a
                                                Pentium 4 processor, on a motherboard with 64 bit slots ( a very big
                                                advantage), large savings there. You can plug in your 64 bit SATA card there.
                                                This board is built on the Serverworks (CSB 6) chipset from Intel..

                                                On top of this the motherboard features,
                                                (1) Gigabit Ethernet interface (10/100/1000)
                                                (1) Onboard IDE
                                                (1) Single channel U160 SCSI controller(64bit/66Mhz) 68 pin supporting Zero
                                                Channel RAID.
                                                (1) builtin VGA

                                                You have almost everything there, moreover, you do not have to pay the hefty
                                                price for an Opteron or a Xeon in this case, moreover, a very stable chipset
                                                based Motherboard.. Should be adequate for a network for around 20 ~ 25
                                                users. (Classified as an entry level server)

                                                If you are going for an Opteron processor, i would suggest going in for a Tyan
                                                motherboard..

                                                Tyan motherboards are bit more expensive ..


                                                Yasir H Alkaar, Operations Manager
                                                ASUS Dubai (Phone): +971-4-2831774 Ext 10
                                                (Fax): +971-4-2831775
                                                (Mobile): +971-50-5506044


                                                For Tyan Motherboards You can contact, Vinod Nair, Business Manager
                                                (Mediasys), They are the regional distributors for Tyan products..

                                                Mediasys(Phone): +971-4-3903069
                                                (Fax): +971-4-3908192
                                                (Mobile): +971-50-7145042

                                                > If not too much can I have your contact numbers and convenient time to
                                                > speak if you don't mind? .

                                                Manu Abraham
                                                Mobile: +971-50-5165769

                                                >
                                                > Regards,
                                                >
                                                > Rajeev
                                                > O50-5247785
                                                >
                                                >
                                                > -----Original Message-----
                                                > From: Manu Abraham [mailto:manu@...]
                                                > Sent: 02 January 2005 20:51
                                                > To: dubailug@yahoogroups.com
                                                > Subject: Re: [dubailug] Sata Drives
                                                >
                                                > On Sun January 2 2005 7:14 pm, Brad Campbell wrote:
                                                > > Manu Abraham wrote:
                                                > > > Oh you misunderstood, i was talking about SCSI hotswap.. Not SATA
                                                > > > hotswap.. I never tried SATA hotswap.. Hotswap i have only on my HA
                                                > > > servers.
                                                > >
                                                > > Fair enough, SCSI requires some tricky termination issues when
                                                > > removing/adding devices to the bus live. Yes other trickery is required.
                                                >
                                                > No, not for SCA SCSI.. I mean Hot swappable ones.. I can demonstrate it
                                                > sometime for you..
                                                >
                                                > > >>Having said that, if you do lose a stripe you are no worse off that
                                                > > >>developing bad sectors on a non-redundant machine. The remainder of the
                                                > > >>data is going to be ok, just that one stripe is going to be dud. You
                                                > > >> can
                                                > > >
                                                > > > I meant when a drive fails.. and have a bad superblock due to some
                                                > > > reason... It need not be a bad sector alone.. Murphy get's lot of his
                                                > > > friends together..
                                                > >
                                                > > Murphy is a bastard!
                                                >
                                                > I agree..
                                                >
                                                > > > That's why for large systems, XFS for linux is considered to be Linux
                                                > > > on steroids.
                                                > >
                                                > > Perhaps. Heard too many stories about XFS data loss. I stick with Ext-3.
                                                > > Certainly not the best choice, but certainly well tested, reliable and
                                                > > not the worst.
                                                >
                                                > Brad, i used to work in an organization with around 60 SGI machines and
                                                > having
                                                > extreme volumes of data.. all CAE/CAM files are really huge. XFS was a
                                                > solution.. i mean at that point of time..
                                                > And was rock solid.. And that was 2.4.
                                                >
                                                > > > For an idling machine that is perfectly alright, Do you know how
                                                >
                                                > software
                                                >
                                                > > > RAID works ? Idle CPU cycles.. But what if the CPU is not idle ? Having
                                                > > > some crunching work .. You loose performance, that's what you pay the
                                                > > > price for Software RAID.
                                                > >
                                                > > That is why I build NAS boxes. The server does all the crunching, the NAS
                                                > > boxes just manage the RAID. It still works out cheaper and more reliable
                                                >
                                                > in
                                                >
                                                > > the long run. On the flip side, you lose write performance, but provided
                                                > > the array is not degraded read performance is nearly as good as RAID-0 as
                                                > > no calculation takes place at all.
                                                >
                                                > Under the circumstances that i had even the network would be congested ...
                                                > Running a PVM cluster with around 140 PIII nodes.
                                                >
                                                > > > Another one is that when the controller is talking to multiple drives,
                                                > > > your PCI bus can easily get flooded, but when a dedicated controller is
                                                > > > talking to the disk, the PCI bus is relatively free, processor also..
                                                > > >
                                                > > > I believe you understand what i really mean..
                                                > >
                                                > > I do.
                                                > >
                                                > > > If you are talking about something like the visualization systems we
                                                >
                                                > had,
                                                >
                                                > > > software RAID is an utter NO .. No .. for the same said reasons ..
                                                > >
                                                > > True, but if you are talking about high performance systems then any form
                                                > > of raid-5 is a no-no.
                                                >
                                                > That is why Hardware RAID is used .. All SGI machines what we had were all
                                                > U320, while the Linux servers were running U160 DPT I2O..
                                                >
                                                > If you understand what I20 is, it gives you lot of performance boost.. In
                                                > the
                                                > sense that all processing is offloaded to the card's processor. This card
                                                > is not very expensive.. It is almost the same price.. Only the SCSI drives
                                                > are more expensive..
                                                >
                                                > > >>If you are after SATA drive cages, Supermicro make some excellent ones
                                                > > >> (and have a local rep who usually has stock in Dubai who is actually
                                                > > >> helpful! I think they stock 3Ware controllers also). I have 2x5 bay
                                                > > >> units in one machine and I have just ordered another 3x5 bay units for
                                                > > >> my NAS box.
                                                > > >
                                                > > > I prefer HA servers to be branded.
                                                > >
                                                > > So do I, but sometimes budget dictates otherwise. I had 1000AED to build
                                                > > the office fileserver, which dictates cheap components and hand
                                                >
                                                > True..
                                                >
                                                > > built/configured machines. It's still RAID-1 and has not so much as
                                                > > farted in the last 18 months. We keep good backups in any case, but
                                                > > still, you can't argue with 1000AED to run a 25 PC network with full
                                                > > mail/fileserver/printserver/nat/firewall/web/gateway services.
                                                > >
                                                > > I'm tired of this conversation. I do it my way, on a budget. You do it
                                                >
                                                > Ok let's stop the conversation here..
                                                >
                                                >
                                                > Manu
                                                >
                                                >
                                                >
                                                > Yahoo! Groups Links
                                                >
                                                >
                                                >
                                                >
                                                >
                                                >
                                                >
                                                >
                                                >
                                                > Yahoo! Groups Links
                                                >
                                                >
                                                >
                                              • Dan Dan the IT Man
                                                A Kenny G Concert. Thats Bad - You must really hate intel!
                                                Message 23 of 29 , Jan 2, 2005
                                                  A Kenny G Concert. Thats Bad - You must really hate intel!




                                                  On Mon, 03 Jan 2005 10:26:39 +0400, Brad Campbell <brad@...> wrote:
                                                  >
                                                  > Rajeev wrote:
                                                  > > Dear Manu and Brad,
                                                  > >
                                                  > > Thanks for your support and sharing your knowledge. It is really
                                                  > > informative.
                                                  > >
                                                  > > Before you stop this conversation could you please give me a part list (with
                                                  > > manufacturer name and model numbers if possible) of the NAS server you had
                                                  > > build with 10 drives. I would like to build a similar one with 6 drives and
                                                  > > increase it later as and when the data increases. Also I thought generally
                                                  > > the hardware price are going down and there is no point in buying too much
                                                  > > hardware for a future use at higher price!
                                                  > >
                                                  >
                                                  > I used an ASUS A7V600 Motherboard
                                                  > Athlon 2600+ Processor
                                                  >
                                                  > 3 x Promise SATA150TX4 4-Port PCI SATA controllers
                                                  > http://www.promise.com/product/product_detail_eng.asp?productId=98&familyId=3
                                                  >
                                                  > 512mb of DDR Ram
                                                  >
                                                  > 12 x Maxtor 7Y250M0 Drives
                                                  > http://www.maxtor.com/portal/site/Maxtor/menuitem.ba88f6d7cf664718376049b291346068/?channelpath=/en_us/Products/ATA%20Hard%20Drives/Midline%20Applications/MaXLine%20Plus%20II&productview=Features
                                                  >
                                                  > 2 x Supermicro CSE-M35T 5 Bay hotswap units
                                                  > http://www.supermicro.com/products/accessories/mobilerack/CSE-M35T-1.cfm
                                                  >
                                                  > The Server http://www.wasp.net.au/~brad/p1000250.jpg
                                                  > The PSU http://www.wasp.net.au/~brad/p1000256.jpg
                                                  >
                                                  > As Manu said. There are quirks with the PCI bridge on the ASUS that cause the odd slow down when
                                                  > reading/writing from multiple controllers at once. I don't have a problem with it, but I would
                                                  > recommend an AMD chipset board (I'm and anti Intel biggot). Or use an Intel board and processor (I'd
                                                  > rather attend a Kenny G concert!)
                                                  >
                                                  > Brad
                                                  >
                                                  >
                                                  > Yahoo! Groups Links
                                                  >
                                                  >
                                                  >
                                                  >
                                                  >
                                                • Manu Abraham
                                                  ... I don t like Intel due to their monopolistic attitude, just like M$, but there is no other alternative for me, i am forced to sit down with it, with a sad
                                                  Message 24 of 29 , Jan 2, 2005
                                                    On Mon January 3 2005 10:26 am, Brad Campbell wrote:
                                                    > Rajeev wrote:
                                                    > > Dear Manu and Brad,
                                                    > >
                                                    > > Thanks for your support and sharing your knowledge. It is really
                                                    > > informative.
                                                    > >
                                                    > > Before you stop this conversation could you please give me a part list
                                                    > > (with manufacturer name and model numbers if possible) of the NAS server
                                                    > > you had build with 10 drives. I would like to build a similar one with 6
                                                    > > drives and increase it later as and when the data increases. Also I
                                                    > > thought generally the hardware price are going down and there is no point
                                                    > > in buying too much hardware for a future use at higher price!
                                                    >
                                                    > I used an ASUS A7V600 Motherboard
                                                    > Athlon 2600+ Processor
                                                    >
                                                    > 3 x Promise SATA150TX4 4-Port PCI SATA controllers
                                                    > http://www.promise.com/product/product_detail_eng.asp?productId=98&familyId
                                                    >=3
                                                    >
                                                    > 512mb of DDR Ram
                                                    >
                                                    > 12 x Maxtor 7Y250M0 Drives
                                                    > http://www.maxtor.com/portal/site/Maxtor/menuitem.ba88f6d7cf664718376049b29
                                                    >1346068/?channelpath=/en_us/Products/ATA%20Hard%20Drives/Midline%20Applicati
                                                    >ons/MaXLine%20Plus%20II&productview=Features
                                                    >
                                                    > 2 x Supermicro CSE-M35T 5 Bay hotswap units
                                                    > http://www.supermicro.com/products/accessories/mobilerack/CSE-M35T-1.cfm
                                                    >
                                                    > The Server http://www.wasp.net.au/~brad/p1000250.jpg
                                                    > The PSU http://www.wasp.net.au/~brad/p1000256.jpg
                                                    >
                                                    > As Manu said. There are quirks with the PCI bridge on the ASUS that cause
                                                    > the odd slow down when reading/writing from multiple controllers at once. I
                                                    > don't have a problem with it, but I would recommend an AMD chipset board
                                                    > (I'm and anti Intel biggot). Or use an Intel board and processor (I'd
                                                    > rather attend a Kenny G concert!)
                                                    I don't like Intel due to their monopolistic attitude, just like M$, but there
                                                    is no other alternative for me, i am forced to sit down with it, with a sad
                                                    face. The other alternative being the Power from IBM, which is not very
                                                    mature under Linux.. and as Luka says Power is best run with AIX, and
                                                    expensive also :-(

                                                    The same goes with the Opteron's also, a good Opteron on a good motherboard, i
                                                    can buy a branded server at that price.

                                                    I think monopoly will stay as a monopoly itself for some more time at
                                                    least. :-(

                                                    Manu
                                                    >
                                                    > Brad
                                                    >
                                                    >
                                                    >
                                                    > Yahoo! Groups Links
                                                    >
                                                    >
                                                    >
                                                  • Brad Campbell
                                                    ... Having said that, I d rather sit through a Kenny G concert than any talk by a Microsoft Rep. But yes, I m not an Intel fan. My only three Intel processors
                                                    Message 25 of 29 , Jan 3, 2005
                                                      Dan Dan the IT Man wrote:
                                                      > A Kenny G Concert. Thats Bad - You must really hate intel!
                                                      >
                                                      >

                                                      Having said that, I'd rather sit through a Kenny G concert than any talk by a Microsoft Rep.
                                                      But yes, I'm not an Intel fan. My only three Intel processors have been my original 8088, A
                                                      Pentium-100 and my current Laptop PIII-M (of which the purchase I had no say in).

                                                      All the rest (and there have been close to 50 others) have been either TI (286-20), AMD (Lots
                                                      starting with a 386DX40), NEC (V20 8086 clone in a laptop), Cyrix (heaps of 686 chips), Zilog (Z80
                                                      Bondwell Laptop and Apple-II Z80 Soft card), MOS Tech (6502 Apple II, C64 and other Embedded
                                                      hardware), Motorola (HC11 In embedded hardware and Motorcycle ECU, 68k Mac and Palm and PPC Mac) and
                                                      last but not least ARM (Prior to the Intel buyout).

                                                      Like I actively avoid the purchase/use of Microsoft Software, I do actively avoid the purchase of
                                                      Intel processors/hardware.

                                                      Hmm interesting when you actually look back over the history of what you buy/use over the years.

                                                      Brad
                                                    • Dan Dan the IT Man
                                                      Try the athlon64. I run athlon64 at home with an ASUS A8V-E mainboard, and it runs fine. The moment SuSE released an Opteron based SLES, I ran out gave a
                                                      Message 26 of 29 , Jan 3, 2005
                                                        Try the athlon64. I run athlon64 at home with an ASUS A8V-E
                                                        mainboard, and it runs fine.

                                                        The moment SuSE released an Opteron based SLES, I ran out gave a
                                                        client an AMD Opteron BOX with 2gb RAM, using an MSI K8Master (I
                                                        think) motherboard! It cost me about 4,800 dhs to build. The client
                                                        didn't mind the clone built PC. I thin that is still ok pricing for a
                                                        box. Granted there was no RAID, just used rsync to back up data
                                                        between different drives.

                                                        dan



                                                        On Mon, 3 Jan 2005 10:59:56 +0400, Manu Abraham <manu@...> wrote:
                                                        >
                                                        > On Mon January 3 2005 10:26 am, Brad Campbell wrote:
                                                        > > Rajeev wrote:
                                                        > > > Dear Manu and Brad,
                                                        > > >
                                                        > > > Thanks for your support and sharing your knowledge. It is really
                                                        > > > informative.
                                                        > > >
                                                        > > > Before you stop this conversation could you please give me a part list
                                                        > > > (with manufacturer name and model numbers if possible) of the NAS server
                                                        > > > you had build with 10 drives. I would like to build a similar one with 6
                                                        > > > drives and increase it later as and when the data increases. Also I
                                                        > > > thought generally the hardware price are going down and there is no point
                                                        > > > in buying too much hardware for a future use at higher price!
                                                        > >
                                                        > > I used an ASUS A7V600 Motherboard
                                                        > > Athlon 2600+ Processor
                                                        > >
                                                        > > 3 x Promise SATA150TX4 4-Port PCI SATA controllers
                                                        > > http://www.promise.com/product/product_detail_eng.asp?productId=98&familyId
                                                        > >=3
                                                        > >
                                                        > > 512mb of DDR Ram
                                                        > >
                                                        > > 12 x Maxtor 7Y250M0 Drives
                                                        > > http://www.maxtor.com/portal/site/Maxtor/menuitem.ba88f6d7cf664718376049b29
                                                        > >1346068/?channelpath=/en_us/Products/ATA%20Hard%20Drives/Midline%20Applicati
                                                        > >ons/MaXLine%20Plus%20II&productview=Features
                                                        > >
                                                        > > 2 x Supermicro CSE-M35T 5 Bay hotswap units
                                                        > > http://www.supermicro.com/products/accessories/mobilerack/CSE-M35T-1.cfm
                                                        > >
                                                        > > The Server http://www.wasp.net.au/~brad/p1000250.jpg
                                                        > > The PSU http://www.wasp.net.au/~brad/p1000256.jpg
                                                        > >
                                                        > > As Manu said. There are quirks with the PCI bridge on the ASUS that cause
                                                        > > the odd slow down when reading/writing from multiple controllers at once. I
                                                        > > don't have a problem with it, but I would recommend an AMD chipset board
                                                        > > (I'm and anti Intel biggot). Or use an Intel board and processor (I'd
                                                        > > rather attend a Kenny G concert!)
                                                        > I don't like Intel due to their monopolistic attitude, just like M$, but there
                                                        > is no other alternative for me, i am forced to sit down with it, with a sad
                                                        > face. The other alternative being the Power from IBM, which is not very
                                                        > mature under Linux.. and as Luka says Power is best run with AIX, and
                                                        > expensive also :-(
                                                        >
                                                        > The same goes with the Opteron's also, a good Opteron on a good motherboard, i
                                                        > can buy a branded server at that price.
                                                        >
                                                        > I think monopoly will stay as a monopoly itself for some more time at
                                                        > least. :-(
                                                        >
                                                        > Manu
                                                        > >
                                                        > > Brad
                                                        > >
                                                        > >
                                                        > >
                                                        > > Yahoo! Groups Links
                                                        > >
                                                        > >
                                                        > >
                                                        >
                                                        > Yahoo! Groups Links
                                                        >
                                                        >
                                                        >
                                                        >
                                                        >
                                                      Your message has been successfully submitted and would be delivered to recipients shortly.