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

7921Re:NSLU2 Drops Connection to LARGE Disk

Expand Messages
  • Mai Kee Reis
    Nov 4, 2009
    • 0 Attachment
      I copied 1 TB of files from one disk to another at the same NSLU2 (new
      266MHz from factory).
      All okay!
      My housings are both IcyBox "IB-351 StUS-B" (one purchased in mid 2008,
      one in October 2009).
      My disks are WD10EACS 1TB SATA (same purchase data).
      OS is SlugOS/LE 5.3 with addition of the optware package, SMB and screen.
      Screen is required to do such long lasting actions from ssh terminal -
      otherwise one has to keep
      the ssh terminal all time open and connected and the ssh guest computer
      all times on.

      The Slug didn't got hot.
      Say 45°C instead of 38°C when idling (estimation related to my own body
      temperature).

      So, if anybody gots problems like dnabble got, he may try my test.
      If this test also fails, the Slug or the USB Disk Housings or the Power
      Supply Units are scrap.

      see:

      MySlug:~# sreen -r
      MySlug:~# dd bs=2048k if=/dev/sdb3 of=/dev/sdc3
      473862+1 records in
      473862+1 records out
      993761846784 bytes (994 GB) copied, 273498 seconds, 3.6 MB/s
      MySlug:~#

      Note:
      273498 sec = 4558 min = 76 h = 3.2 days!

      Okay, pretty slow. But it runs hell stable!
      Without the 2MB buffer, it would be much slower ;-)

      with best regards,
      Mai Kee

      > There is 1 message in this issue.
      >
      > Topics in this digest:
      >
      > 1a. Re: NSLU2 Drops Connection to LARGE Disk
      > From: dnabble
      >
      >
      > Message
      > ________________________________________________________________________
      > 1a. Re: NSLU2 Drops Connection to LARGE Disk
      > Posted by: "dnabble" david.nelson.wi@... cnczane
      > Date: Wed Oct 28, 2009 4:45 am ((PDT))
      >
      >
      > I bought a "new" (but alas discontinued) NSLU2, a pair of 1.5T internal eSATA
      > drives, and a pair of "almost" matching BlackX HDD USB/eSATA "docking
      > stations."
      >
      > Not so prominently printed on the BlackX boxes, is the phrase: "SUPPORT UP
      > TO 1TB (1,000GB) H.D.D."
      >
      > Rats!!
      >
      > Nevertheless, I decided to try, and went ahead with formatting, figuring to
      > see an error message, signaling failure.
      >
      > Formats appeared to succeed, as did unslinging (SlugOS/LE), as did rsyncing,
      > for a time, and therein lies a tale.
      >
      > As the drive filled up, the rsync process would hang inexplicably, and hang
      > the slug: all and any terminal sessions would become unresponsive and new
      > ssh requests ignored.
      >
      > Reboots always solved the problem, and always the slug was good for whatever
      > you wanted, which seems to suggest overheating was not the problem. In fact
      > the hanging was always associated with an rsync copying more files inbound.
      >
      > I was worried at first that, for the first time (in 4 slugs), the
      > over-clocking to 266 MHz had bit me, but (see overheating consideration
      > above) the hanging should have been random and not so connected with rsync.
      >
      > The mystifier was that (1) rsync would hang during a large (1GB) file
      > transfer, (2) reboot, (3) rsync might appear to complete that file transfer
      > and hang during an ensuing one. The disk "fill" was at around 80% as I
      > recall.
      >
      > The problem also MIGHT have been related to my use of a 32MB "swapfile"
      > instead of a larger, and real-er, swap "partition." (SlugOS/LE is smart
      > enough managing memory to be able to get by with a smallish swapfile, in my
      > understanding. [See "Install and Turnup SlugOS" at www.nslu2-linux.org]
      >
      > At any rate, my solution:
      > Decide to live within the means of the BlackX docking stations and partition
      > the drive with TWO sizable data partitions (91000 blocks) and a THIRD
      > remainder swap partition (400 blocks).
      >
      > So far, I filled the first data partition to 82%, and have started the
      > second. The acid test will come when that 2nd data partition hits around
      > 80-90%, but the NSLU2 hasn't so much as burped. IF CPU or disk were
      > overheating, or IF it was over-overclocked, there should've been a burp,
      > because the rsync stress is identical to the first case.
      >
      > Cheers!
      >
      >
      >
    • Show all 14 messages in this topic