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

Re: [irlp-embedded] aprs_status / perl?

Expand Messages
  • Randy Hammock
    ... I ll be happy to assist in any way. I an make what ever mods are needed to the aprs_status bash script to work with the new system. I need to do some more
    Message 1 of 27 , Apr 10, 2007
    • 0 Attachment
      On Tue, 2007-04-10 at 08:22 -0700, David Cameron - IRLP wrote:
      > From what I can tell, all that is sent to the APRS network is two
      > lines
      > of text. Since this is the case, there is already a nice tool to do
      > this.

      I'll be happy to assist in any way. I an make what ever mods are needed
      to the aprs_status bash script to work with the new system. I need to do
      some more testing here now. The APRS2 net has made some changes in the
      last month. Don't know if I need to make changes to work with the new
      structure. I was unable to send out any status packet this morning when
      I did some testing.

      Dave'e requested info was sent off-line earlier.

      --
      Randy Hammock KC6HUR
      IRLP 4494 EchoLink 120688 (KC6HUR-L)
      http://kc6hur.net
      http://irlp.kc6hur.net
    • Keith VE7GDH
      Randy KC6HUR wrote... ... I upgraded from the APRS status script from Alan ZL1AMW to your ver 3 script yesterday on node 1147. It seems to be working OK. There
      Message 2 of 27 , Apr 10, 2007
      • 0 Attachment
        Randy KC6HUR wrote...

        > I was unable to send out any status packet this morning when I did some
        > testing.

        I upgraded from the APRS status script from Alan ZL1AMW to your ver 3 script
        yesterday on node 1147. It seems to be working OK. There were some small
        differences, such as the hyphen in the object name... IRLP-1147 instead of
        IRLP1147, and IDLE instead of (I think) Link Clear and so on, but it
        appeared to be functioning OK when I connected to it and disconnected.

        I see that I'm replying to the embedded list. Node 1147 is non-embedded.

        Keith VE7GDH
        Node 1148
      • ve7ltd
        I have provided Randy with a working solution that does not use perl. He should have an updated version 3.01 APRS-STATUS script ready soon. The fix was really
        Message 3 of 27 , Apr 10, 2007
        • 0 Attachment
          I have provided Randy with a working solution that does not use
          perl. He should have an updated version 3.01 APRS-STATUS script
          ready soon. The fix was really easy, and uses programs already
          available on the embedded and all IRLP nodes.

          Dave Cameron - VE7LTD

          --- In irlp-embedded@yahoogroups.com, Randy Hammock <rhammock@...>
          wrote:
          >
          > On Tue, 2007-04-10 at 08:22 -0700, David Cameron - IRLP wrote:
          > > From what I can tell, all that is sent to the APRS network is two
          > > lines
          > > of text. Since this is the case, there is already a nice tool to
          do
          > > this.
          >
          > I'll be happy to assist in any way. I an make what ever mods are
          needed
          > to the aprs_status bash script to work with the new system. I need
          to do
          > some more testing here now. The APRS2 net has made some changes in
          the
          > last month. Don't know if I need to make changes to work with the
          new
          > structure. I was unable to send out any status packet this morning
          when
          > I did some testing.
          >
          > Dave'e requested info was sent off-line earlier.
          >
          > --
          > Randy Hammock KC6HUR
          > IRLP 4494 EchoLink 120688 (KC6HUR-L)
          > http://kc6hur.net
          > http://irlp.kc6hur.net
          >
        • David McAnally
          ... I was experimenting with using the netcat (nc) command. Based on what I got back from verbose mode, NC wants to have matching addresses on DNS
          Message 4 of 27 , Apr 10, 2007
          • 0 Attachment
            On 4/10/07, ve7ltd <dcameron@...> wrote:
            I have provided Randy with a working solution that does not use
            perl. He should have an updated version 3.01 APRS-STATUS script
            ready soon. The fix was really easy, and uses programs already
            available on the embedded and all IRLP nodes.

            Dave Cameron - VE7LTD

            I was experimenting with using the netcat (nc) command.  Based on what I got back from verbose mode, NC wants to have matching addresses on DNS reverse/forward lookups. I didn't find an option to disable that.  It seemed to skip non-matching reverse lookups on IP addresses from the aprs DNS hosts list. 

            You probably came up with something better, but this is what "seems" to work for me.  I used the -n option and loop through the IP address list from a host lookup.  I replaced the call to the aprs_status_tx.pl script, with a call to this bash function, with the function added to the top of the aprs_status script.

            (...)

            [ -z "$SCRIPT" ] && SCRIPT=/home/irlp/scripts
            source $SCRIPT/common-functions.sh
            function aprs_status_tx {
                    # Transmit to aprs server:
                    server=" rotate.aprs2.net"
                    port="14580"
                    call="$1"
                    passwd="$2"
                    beacon="$3"
                    logon="User $call pass $passwd vers IRLP-interface 1\n"
                    for HOST in `/usr/bin/host -t a rotate.aprs2.net|cut -d' ' -f4`  ; do
                            echo "$HOST"
                            echo -ne "$logon$call$beacon\n" | $NETCAT -w2 -v -n $HOST $port
                            if [ "$?" -eq "0" ] ; then
                                    break
                            fi
                    done
            }

            (...)

            aprs_status_tx "${APRS_CALL}" "${APRS_VALIDATION}" "${BTEXT}"

            --
            Regards,
            David McAnally, WD5M
          • Randy Hammock
            ... I basically decided to just let nc do it s thing without scraping the host list. Maybe I ll modify it later for version 4.1. ... -- Randy Hammock KC6HUR
            Message 5 of 27 , Apr 10, 2007
            • 0 Attachment
              On Tue, 2007-04-10 at 23:50 -0500, David McAnally wrote:
              > On 4/10/07, ve7ltd <dcameron@...> wrote:
              >
              > I have provided Randy with a working solution that does not
              > use
              > perl. He should have an updated version 3.01 APRS-STATUS
              > script
              > ready soon. The fix was really easy, and uses programs already
              > available on the embedded and all IRLP nodes.
              >
              > Dave Cameron - VE7LTD
              >
              > I was experimenting with using the netcat (nc) command. Based on what
              > I got back from verbose mode, NC wants to have matching addresses on
              > DNS reverse/forward lookups. I didn't find an option to disable that.
              > It seemed to skip non-matching reverse lookups on IP addresses from
              > the aprs DNS hosts list.
              >
              > You probably came up with something better, but this is what "seems"
              > to work for me. I used the -n option and loop through the IP address
              > list from a host lookup. I replaced the call to the aprs_status_tx.pl
              > script, with a call to this bash function, with the function added to
              > the top of the aprs_status script.
              >
              > (...)
              >
              > [ -z "$SCRIPT" ] && SCRIPT=/home/irlp/scripts
              > source $SCRIPT/common-functions.sh
              > function aprs_status_tx {
              > # Transmit to aprs server:
              > server="rotate.aprs2.net"
              > port="14580"
              > call="$1"
              > passwd="$2"
              > beacon="$3"
              > logon="User $call pass $passwd vers IRLP-interface 1\n"
              > for HOST in `/usr/bin/host -t a rotate.aprs2.net|cut -d' ' -
              > f4` ; do
              > echo "$HOST"
              > echo -ne "$logon$call$beacon\n" | $NETCAT -w2 -v -n
              > $HOST $port
              > if [ "$?" -eq "0" ] ; then
              > break
              > fi
              > done
              > }
              >
              > (...)
              >
              > aprs_status_tx "${APRS_CALL}" "${APRS_VALIDATION}" "${BTEXT}"

              I basically decided to just let nc do it's thing without scraping the
              host list. Maybe I'll modify it later for version 4.1.

              >
              > --
              > Regards,
              > David McAnally, WD5M
              >
              >
              >
              --
              Randy Hammock KC6HUR
              IRLP 4494 EchoLink 120688 (KC6HUR-L)
              http://kc6hur.net
              http://irlp.kc6hur.net
            • Randy Hammock
              ... I released 4.1 which incorporates the DNS preprocessing. I only used the for loop and changed a few vars to match what I used. -- Randy Hammock KC6HUR
              Message 6 of 27 , Apr 10, 2007
              • 0 Attachment
                On Tue, 2007-04-10 at 23:50 -0500, David McAnally wrote:
                > On 4/10/07, ve7ltd <dcameron@...> wrote:
                >
                > I have provided Randy with a working solution that does not
                > use
                > perl. He should have an updated version 3.01 APRS-STATUS
                > script
                > ready soon. The fix was really easy, and uses programs already
                > available on the embedded and all IRLP nodes.
                >
                > Dave Cameron - VE7LTD
                >
                > I was experimenting with using the netcat (nc) command. Based on what
                > I got back from verbose mode, NC wants to have matching addresses on
                > DNS reverse/forward lookups. I didn't find an option to disable that.
                > It seemed to skip non-matching reverse lookups on IP addresses from
                > the aprs DNS hosts list.
                >
                > You probably came up with something better, but this is what "seems"
                > to work for me. I used the -n option and loop through the IP address
                > list from a host lookup. I replaced the call to the aprs_status_tx.pl
                > script, with a call to this bash function, with the function added to
                > the top of the aprs_status script.
                >
                > (...)
                >
                > [ -z "$SCRIPT" ] && SCRIPT=/home/irlp/scripts
                > source $SCRIPT/common-functions.sh
                > function aprs_status_tx {
                > # Transmit to aprs server:
                > server="rotate.aprs2.net"
                > port="14580"
                > call="$1"
                > passwd="$2"
                > beacon="$3"
                > logon="User $call pass $passwd vers IRLP-interface 1\n"
                > for HOST in `/usr/bin/host -t a rotate.aprs2.net|cut -d' ' -
                > f4` ; do
                > echo "$HOST"
                > echo -ne "$logon$call$beacon\n" | $NETCAT -w2 -v -n
                > $HOST $port
                > if [ "$?" -eq "0" ] ; then
                > break
                > fi
                > done
                > }
                >
                > (...)
                >
                > aprs_status_tx "${APRS_CALL}" "${APRS_VALIDATION}" "${BTEXT}"

                I released 4.1 which incorporates the DNS preprocessing. I only used the
                "for" loop and changed a few vars to match what I used.

                --
                Randy Hammock KC6HUR
                IRLP 4494 EchoLink 120688 (KC6HUR-L)
                http://kc6hur.net
                http://irlp.kc6hur.net
              • Keith VE7GDH
                Randy KC6HUR wrote... ... I had ver 4.0 OK but I couldn t get 4.1 to work so I went back to 4.0 until I can get it sorted out. Keith VE7GDH Node 1148
                Message 7 of 27 , Apr 11, 2007
                • 0 Attachment
                  Randy KC6HUR wrote...

                  > I released 4.1 which incorporates the DNS preprocessing. I only
                  > used the "for" loop and changed a few vars to match what I used.

                  I had ver 4.0 OK but I couldn't get 4.1 to work so I went back to 4.0 until
                  I can get it sorted out.

                  Keith VE7GDH
                  Node 1148
                • dcameron@irlp.net
                  The program host on embedded nodes is not standard, and this will not work. I don t understand the problem though. What is netcat not doing right? Dave Sent
                  Message 8 of 27 , Apr 11, 2007
                  • 0 Attachment
                    The program "host" on embedded nodes is not standard, and this will not work. I don't understand the problem though. What is netcat not doing right?

                    Dave
                    Sent from my BlackBerry device on the Rogers Wireless Network

                    -----Original Message-----
                    From: Randy Hammock <rhammock@...>
                    Date: Tue, 10 Apr 2007 23:23:58
                    To:irlp-embedded@yahoogroups.com
                    Subject: Re: [irlp-embedded] Re: aprs_status / perl?

                    On Tue, 2007-04-10 at 23:50 -0500, David McAnally wrote:
                    > On 4/10/07, ve7ltd <dcameron@irlp.: <mailto:dcameron%40irlp.net> net> wrote:
                    >
                    > I have provided Randy with a working solution that does not
                    > use
                    > perl. He should have an updated version 3.01 APRS-STATUS
                    > script
                    > ready soon. The fix was really easy, and uses programs already
                    > available on the embedded and all IRLP nodes.
                    >
                    > Dave Cameron - VE7LTD
                    >
                    > I was experimenting with using the netcat (nc) command. Based on what
                    > I got back from verbose mode, NC wants to have matching addresses on
                    > DNS reverse/forward lookups. I didn't find an option to disable that.
                    > It seemed to skip non-matching reverse lookups on IP addresses from
                    > the aprs DNS hosts list.
                    >
                    > You probably came up with something better, but this is what "seems"
                    > to work for me. I used the -n option and loop through the IP address
                    > list from a host lookup. I replaced the call to the aprs_status_tx.pl
                    > script, with a call to this bash function, with the function added to
                    > the top of the aprs_status script.
                    >
                    > (...)
                    >
                    > [ -z "$SCRIPT" ] && SCRIPT=/home/irlp/scripts
                    > source $SCRIPT/common-functions.sh
                    > function aprs_status_tx {
                    > # Transmit to aprs server:
                    > server="rotate.aprs2.net"
                    > port="14580"
                    > call="$1"
                    > passwd="$2"
                    > beacon="$3"
                    > logon="User $call pass $passwd vers IRLP-interface 1\n"
                    > for HOST in `/usr/bin/host -t a rotate.aprs2.net|cut -d' ' -
                    > f4` ; do
                    > echo "$HOST"
                    > echo -ne "$logon$call$beacon\n" | $NETCAT -w2 -v -n
                    > $HOST $port
                    > if [ "$?" -eq "0" ] ; then
                    > break
                    > fi
                    > done
                    > }
                    >
                    > (...)
                    >
                    > aprs_status_tx "${APRS_CALL}" "${APRS_VALIDATION}" "${BTEXT}"

                    I released 4.1 which incorporates the DNS preprocessing. I only used the
                    "for" loop and changed a few vars to match what I used.

                    --
                    Randy Hammock KC6HUR
                    IRLP 4494 EchoLink 120688 (KC6HUR-L)
                    http://kc6hur.: <http://kc6hur.net> net
                    http://irlp.: <http://irlp.kc6hur.net> kc6hur.net
                  • Randy Hammock
                    netcat apparently does reverse DNS. Seems most of the APRS-IS machines are on DSL/Cable connection. netcat does the DNS lookup to get the IP then it does an
                    Message 9 of 27 , Apr 11, 2007
                    • 0 Attachment
                      netcat apparently does reverse DNS. Seems most of the APRS-IS machines
                      are on DSL/Cable connection. netcat does the DNS lookup to get the IP
                      then it does an inverse lookup to verify the name and gets the "real"
                      hostname from the ISP rather than the user registered name for the
                      address. When it sees the hostname difference, it skips to the next
                      entry in the rotate list. Skips as many 6 or more servers before finding
                      a match. The host preprocessing Dave FLY came with allows passing just
                      the IP to netcat which when in IP number mode (-n) accepts the number
                      with doing any DNS lookups.

                      Well, version 4.0 is still up on my server, and can be downloaded. Just
                      change the 4.1 in the download instructions to 4.0.
                      On Wed, 2007-04-11 at 13:17 +0000, dcameron@... wrote:
                      > The program "host" on embedded nodes is not standard, and this will not work. I don't understand the problem though. What is netcat not doing right?
                      >
                      > Dave
                      > Sent from my BlackBerry device on the Rogers Wireless Network
                      >
                      > -----Original Message-----
                      > From: Randy Hammock <rhammock@...>
                      > Date: Tue, 10 Apr 2007 23:23:58
                      > To:irlp-embedded@yahoogroups.com
                      > Subject: Re: [irlp-embedded] Re: aprs_status / perl?
                      >
                      > On Tue, 2007-04-10 at 23:50 -0500, David McAnally wrote:
                      > > On 4/10/07, ve7ltd <dcameron@irlp.: <mailto:dcameron%40irlp.net> net> wrote:
                      > >
                      > > I have provided Randy with a working solution that does not
                      > > use
                      > > perl. He should have an updated version 3.01 APRS-STATUS
                      > > script
                      > > ready soon. The fix was really easy, and uses programs already
                      > > available on the embedded and all IRLP nodes.
                      > >
                      > > Dave Cameron - VE7LTD
                      > >
                      > > I was experimenting with using the netcat (nc) command. Based on what
                      > > I got back from verbose mode, NC wants to have matching addresses on
                      > > DNS reverse/forward lookups. I didn't find an option to disable that.
                      > > It seemed to skip non-matching reverse lookups on IP addresses from
                      > > the aprs DNS hosts list.
                      > >
                      > > You probably came up with something better, but this is what "seems"
                      > > to work for me. I used the -n option and loop through the IP address
                      > > list from a host lookup. I replaced the call to the aprs_status_tx.pl
                      > > script, with a call to this bash function, with the function added to
                      > > the top of the aprs_status script.
                      > >
                      > > (...)
                      > >
                      > > [ -z "$SCRIPT" ] && SCRIPT=/home/irlp/scripts
                      > > source $SCRIPT/common-functions.sh
                      > > function aprs_status_tx {
                      > > # Transmit to aprs server:
                      > > server="rotate.aprs2.net"
                      > > port="14580"
                      > > call="$1"
                      > > passwd="$2"
                      > > beacon="$3"
                      > > logon="User $call pass $passwd vers IRLP-interface 1\n"
                      > > for HOST in `/usr/bin/host -t a rotate.aprs2.net|cut -d' ' -
                      > > f4` ; do
                      > > echo "$HOST"
                      > > echo -ne "$logon$call$beacon\n" | $NETCAT -w2 -v -n
                      > > $HOST $port
                      > > if [ "$?" -eq "0" ] ; then
                      > > break
                      > > fi
                      > > done
                      > > }
                      > >
                      > > (...)
                      > >
                      > > aprs_status_tx "${APRS_CALL}" "${APRS_VALIDATION}" "${BTEXT}"
                      >
                      > I released 4.1 which incorporates the DNS preprocessing. I only used the
                      > "for" loop and changed a few vars to match what I used.
                      >
                      > --
                      > Randy Hammock KC6HUR
                      > IRLP 4494 EchoLink 120688 (KC6HUR-L)
                      > http://kc6hur.: <http://kc6hur.net> net
                      > http://irlp.: <http://irlp.kc6hur.net> kc6hur.net
                      >
                      >
                      >
                      >
                      >
                      > Yahoo! Groups Links
                      >
                      >
                      >
                      --
                      Randy Hammock KC6HUR
                      IRLP 4494 EchoLink 120688 (KC6HUR-L)
                      http://kc6hur.net
                      http://irlp.kc6hur.net
                    • David McAnally
                      ... I m beginning to wonder of those verbose messages are just warning messages. The 0.7.1 GNU version of NC has continue; statements for similar messages,
                      Message 10 of 27 , Apr 11, 2007
                      • 0 Attachment
                        On 4/11/07, Randy Hammock <rhammock@...> wrote:
                        netcat apparently does reverse DNS. Seems most of the APRS-IS machines
                        are on DSL/Cable connection. netcat does the DNS lookup to get the IP
                        then it does an inverse lookup to verify the name and gets the "real"
                        hostname from the ISP rather than the user registered name for the
                        address. When it sees the hostname difference, it skips to the next
                        entry in the rotate list. Skips as many 6 or more servers before finding
                        a match. The host preprocessing Dave FLY came with allows passing just
                        the IP to netcat which when in IP number mode (-n) accepts the number
                        with doing any DNS lookups.

                        Well, version 4.0 is still up on my server, and can be downloaded. Just
                        change the 4.1 in the download instructions to 4.0.
                        On Wed, 2007-04-11 at 13:17 +0000, dcameron@... wrote:
                        > The program "host" on embedded nodes is not standard, and this will not work. I don't understand the problem though. What is netcat not doing right?
                        >
                        > Dave
                        > Sent from my BlackBerry device on the Rogers Wireless Network

                        I'm beginning to wonder of those verbose messages are just warning messages.  The 0.7.1 GNU version of NC has "continue;" statements for similar messages, but I didn't find the exact same text in that code. 

                        Dave, can you just check the src of NC you use, in network.c and see if this is just a warning message displayed with verbose options?  If so, we can drop the host IP loop method.

                        --
                        Regards,
                        David McAnally, WD5M
                      • Randy Hammock
                        ... one where it can do an inverse lookup, does it s thing, then it quits. That s what I saw when using the -v -v option for very verbose output. ... -- Randy
                        Message 11 of 27 , Apr 11, 2007
                        • 0 Attachment
                          On Wed, 2007-04-11 at 14:33 -0500, David McAnally wrote:
                          > On 4/11/07, Randy Hammock <rhammock@...> wrote:
                          >
                          > netcat apparently does reverse DNS. Seems most of the APRS-IS
                          > machines
                          > are on DSL/Cable connection. netcat does the DNS lookup to get
                          > the IP
                          > then it does an inverse lookup to verify the name and gets the
                          > "real"
                          > hostname from the ISP rather than the user registered name for
                          > the
                          > address. When it sees the hostname difference, it skips to the
                          > next
                          > entry in the rotate list. Skips as many 6 or more servers
                          > before finding
                          > a match. The host preprocessing Dave FLY came with allows
                          > passing just
                          > the IP to netcat which when in IP number mode (-n) accepts the
                          > number
                          > with doing any DNS lookups.
                          >
                          > Well, version 4.0 is still up on my server, and can be
                          > downloaded. Just
                          > change the 4.1 in the download instructions to 4.0.
                          > On Wed, 2007-04-11 at 13:17 +0000, dcameron@... wrote:
                          > > The program "host" on embedded nodes is not standard, and
                          > this will not work. I don't understand the problem though.
                          > What is netcat not doing right?
                          > >
                          > > Dave
                          > > Sent from my BlackBerry device on the Rogers Wireless
                          > Network
                          >
                          > I'm beginning to wonder of those verbose messages are just warning
                          > messages. The 0.7.1 GNU version of NC has "continue;" statements for
                          > similar messages, but I didn't find the exact same text in that code.

                          >From what I can tell, nc merrily skips through the list until it finds
                          one where it can do an inverse lookup, does it's thing, then it quits.
                          That's what I saw when using the -v -v option for very verbose output.

                          >
                          > Dave, can you just check the src of NC you use, in network.c and see
                          > if this is just a warning message displayed with verbose options? If
                          > so, we can drop the host IP loop method.
                          >
                          > --
                          > Regards,
                          > David McAnally, WD5M

                          --
                          Randy Hammock KC6HUR
                          IRLP 4494 EchoLink 120688 (KC6HUR-L)
                          http://kc6hur.net
                          http://irlp.kc6hur.net
                        • David McAnally
                          ... That is what I thought initially. But when you look up the IP address that nc ends up using, it seems to be the first of those mismatched host names listed
                          Message 12 of 27 , Apr 11, 2007
                          • 0 Attachment
                            On 4/11/07, Randy Hammock <rhammock@...> wrote:
                            >From what I can tell, nc merrily skips through the list until it finds
                            one where it can do an inverse lookup, does it's thing, then it quits.
                            That's what I saw when using the -v -v option for very verbose output.


                            That is what I thought initially. But when you look up the IP address that nc ends up using, it seems to be the first of those mismatched host names listed from nc.  I think nc is just iterating through the list doing reverse lookups on all of them before it actually tries to connect.  And then it seems to use the first one in the list. 

                            Since the first attempt usually connects, I haven't seen it try more than one address yet.  Guess I could block some addresses with iptables and see what happens.  I  did run it enough to see it get a different order of addresses from the list at least once, and so connected to a different server.

                            I see that nc closes the connection with a "net timeout" after sending the message.  Doesn't aprs igate syntax have a "quit" message/command to close the connection?  Maybe not.

                            Oh well, we've probably filled up enough mailboxes with this off-topic discussion.  I'd suggest we just drop the host lookup loop, until we find out for sure it is needed.

                            --
                            Regards,
                            David McAnally, WD5M
                          • dcameron@irlp.net
                            The script uses the installed version of nc, so there is no default version. The host program is all screwy on the embedded nodes. We will have to find a
                            Message 13 of 27 , Apr 11, 2007
                            • 0 Attachment
                              The script uses the installed version of nc, so there is no default version. The host program is all screwy on the embedded nodes. We will have to find a better solution, or compile something new for the embedded nodes. All embedded nodes have the IS_EMBEDDED flag set to YES.

                              Dave Cameron
                              Sent from my BlackBerry device on the Rogers Wireless Network

                              -----Original Message-----
                              From: Randy Hammock <rhammock@...>
                              Date: Wed, 11 Apr 2007 17:59:55
                              To:irlp-embedded@yahoogroups.com
                              Subject: Re: [irlp-embedded] Re: aprs_status / perl?

                              On Wed, 2007-04-11 at 14:33 -0500, David McAnally wrote:
                              > On 4/11/07, Randy Hammock <rhammock@ca.: <mailto:rhammock%40ca.rr.com> rr.com> wrote:
                              >
                              > netcat apparently does reverse DNS. Seems most of the APRS-IS
                              > machines
                              > are on DSL/Cable connection. netcat does the DNS lookup to get
                              > the IP
                              > then it does an inverse lookup to verify the name and gets the
                              > "real"
                              > hostname from the ISP rather than the user registered name for
                              > the
                              > address. When it sees the hostname difference, it skips to the
                              > next
                              > entry in the rotate list. Skips as many 6 or more servers
                              > before finding
                              > a match. The host preprocessing Dave FLY came with allows
                              > passing just
                              > the IP to netcat which when in IP number mode (-n) accepts the
                              > number
                              > with doing any DNS lookups.
                              >
                              > Well, version 4.0 is still up on my server, and can be
                              > downloaded. Just
                              > change the 4.1 in the download instructions to 4.0.
                              > On Wed, 2007-04-11 at 13:17 +0000, dcameron@irlp.: <mailto:dcameron%40irlp.net> net wrote:
                              > > The program "host" on embedded nodes is not standard, and
                              > this will not work. I don't understand the problem though.
                              > What is netcat not doing right?
                              > >
                              > > Dave
                              > > Sent from my BlackBerry device on the Rogers Wireless
                              > Network
                              >
                              > I'm beginning to wonder of those verbose messages are just warning
                              > messages. The 0.7.1 GNU version of NC has "continue;" statements for
                              > similar messages, but I didn't find the exact same text in that code.

                              >From what I can tell, nc merrily skips through the list until it finds
                              one where it can do an inverse lookup, does it's thing, then it quits.
                              That's what I saw when using the -v -v option for very verbose output.

                              >
                              > Dave, can you just check the src of NC you use, in network.c and see
                              > if this is just a warning message displayed with verbose options? If
                              > so, we can drop the host IP loop method.
                              >
                              > --
                              > Regards,
                              > David McAnally, WD5M

                              --
                              Randy Hammock KC6HUR
                              IRLP 4494 EchoLink 120688 (KC6HUR-L)
                              http://kc6hur.: <http://kc6hur.net> net
                              http://irlp.: <http://irlp.kc6hur.net> kc6hur.net
                            • Randy Hammock
                              ... I just rolled my site back to the 4.0 version with the host preprocessing. This will work for all nodes. Those who wish to experiment on their own, feel
                              Message 14 of 27 , Apr 11, 2007
                              • 0 Attachment
                                On Thu, 2007-04-12 at 02:19 +0000, dcameron@... wrote:
                                > The script uses the installed version of nc, so there is no default version. The host program is all screwy on the embedded nodes. We will have to find a better solution, or compile something new for the embedded nodes. All embedded nodes have the IS_EMBEDDED flag set to YES.

                                I just rolled my site back to the 4.0 version with the host
                                preprocessing. This will work for all nodes. Those who wish to
                                experiment on their own, feel free to do so.

                                > Dave Cameron
                                > Sent from my BlackBerry device on the Rogers Wireless Network
                                >
                                > -----Original Message-----
                                > From: Randy Hammock <rhammock@...>
                                > Date: Wed, 11 Apr 2007 17:59:55
                                > To:irlp-embedded@yahoogroups.com
                                > Subject: Re: [irlp-embedded] Re: aprs_status / perl?
                                >
                                > On Wed, 2007-04-11 at 14:33 -0500, David McAnally wrote:
                                > > On 4/11/07, Randy Hammock <rhammock@ca.: <mailto:rhammock%40ca.rr.com> rr.com> wrote:
                                > >
                                > > netcat apparently does reverse DNS. Seems most of the APRS-IS
                                > > machines
                                > > are on DSL/Cable connection. netcat does the DNS lookup to get
                                > > the IP
                                > > then it does an inverse lookup to verify the name and gets the
                                > > "real"
                                > > hostname from the ISP rather than the user registered name for
                                > > the
                                > > address. When it sees the hostname difference, it skips to the
                                > > next
                                > > entry in the rotate list. Skips as many 6 or more servers
                                > > before finding
                                > > a match. The host preprocessing Dave FLY came with allows
                                > > passing just
                                > > the IP to netcat which when in IP number mode (-n) accepts the
                                > > number
                                > > with doing any DNS lookups.
                                > >
                                > > Well, version 4.0 is still up on my server, and can be
                                > > downloaded. Just
                                > > change the 4.1 in the download instructions to 4.0.
                                > > On Wed, 2007-04-11 at 13:17 +0000, dcameron@irlp.: <mailto:dcameron%40irlp.net> net wrote:
                                > > > The program "host" on embedded nodes is not standard, and
                                > > this will not work. I don't understand the problem though.
                                > > What is netcat not doing right?
                                > > >
                                > > > Dave
                                > > > Sent from my BlackBerry device on the Rogers Wireless
                                > > Network
                                > >
                                > > I'm beginning to wonder of those verbose messages are just warning
                                > > messages. The 0.7.1 GNU version of NC has "continue;" statements for
                                > > similar messages, but I didn't find the exact same text in that code.
                                >
                                > >From what I can tell, nc merrily skips through the list until it finds
                                > one where it can do an inverse lookup, does it's thing, then it quits.
                                > That's what I saw when using the -v -v option for very verbose output.
                                >
                                > >
                                > > Dave, can you just check the src of NC you use, in network.c and see
                                > > if this is just a warning message displayed with verbose options? If
                                > > so, we can drop the host IP loop method.

                                --
                                Randy Hammock KC6HUR
                                IRLP 4494 EchoLink 120688 (KC6HUR-L)
                                http://kc6hur.net
                                http://irlp.kc6hur.net
                              • David McAnally
                                ... Ok, I thought I was done. But, after trying this new method on some more nodes, I find that the NC may not disconnect, and may hang the script on some
                                Message 15 of 27 , Apr 11, 2007
                                • 0 Attachment
                                  On 4/11/07, dcameron@... <dcameron@...> wrote:
                                  The script uses the installed version of nc, so there is no default version. The host program is all screwy on the embedded nodes. We will have to find a better solution, or compile something new for the embedded nodes. All embedded nodes have the IS_EMBEDDED flag set to YES.


                                  Ok, I thought I was done.  But, after trying this new method on some more nodes, I find that the NC may not disconnect, and may hang the script on some nodes.  I have one node on FC5 that does not disconnect after sending the aprs message, even with -w1 timeout flag.  The aprs server sends back time stamp messages periodically, but nc never disconnects.

                                  That node has a GNU NC version 0.7.1.  Interestingly, that nc version also includes a "-c" option to cause it to close the connection on EOF from stdin.  And that fixes the issue, on that node. 

                                  Problem is, other nodes may not have that version of nc.  The version in IRLP doesn't have the -c option.  My RH7.3 nodes have nc version 0.6.1, and also don't have the -c option.

                                  I suspect we will also need to touch a flag or status file in this script, and code to block attempts to run multiple instances of the script at the same time if the status file exists.  Just in case the script hangs. 

                                  --
                                  Regards,
                                  David McAnally, WD5M
                                • David McAnally
                                  ... I compiled GNU NETCAT 0.7.1 on my RH7.3 node. With the new -c disconnect option, that seems to work best. I m about to try it on RH9. It is on
                                  Message 16 of 27 , Apr 11, 2007
                                  • 0 Attachment
                                    On 4/11/07, Randy Hammock <rhammock@...> wrote:
                                    On Thu, 2007-04-12 at 02:19 +0000, dcameron@... wrote:
                                    > The script uses the installed version of nc, so there is no default version. The host program is all screwy on the embedded nodes. We will have to find a better solution, or compile something new for the embedded nodes. All embedded nodes have the IS_EMBEDDED flag set to YES.

                                    I just rolled my site back to the 4.0 version with the host
                                    preprocessing. This will work for all nodes. Those who wish to
                                    experiment on their own, feel free to do so.


                                    I compiled GNU NETCAT 0.7.1 on my RH7.3 node.  With the new -c disconnect option, that seems to work best.  I'm about to try it on RH9.  It is on  sourceforge at <http://netcat.sourceforge.net/ >.

                                    --
                                    Regards,
                                    David McAnally, WD5M
                                  • David McAnally
                                    I don t know why, but one of my RH7.3 nodes would not send the APRS packet using NC 0.7.1. The NC program seems to work, but no update appears to be received
                                    Message 17 of 27 , Apr 12, 2007
                                    • 0 Attachment
                                      I don't know why, but one of my RH7.3 nodes would not send the APRS packet using NC 0.7.1.  The NC program seems to work, but no update appears to be received on APRS igate.  Verbose debugging didn't provide any clues either.

                                      So I left that node running with the PERL for version of aprs_status, until I remembered another possible method.  I don't know if this will end up being the ultimate solution, but this is what I'm using now.

                                      echo -e "$logon$call$beacon" >/dev/tcp/${server}/${port}

                                      Always surprising what you find built-in to BASH and other shells.  No need for NC or PERL. 

                                      I'm just not sure about timeouts on the connection attempts, and the bash manpage is a bit thin on this feature.  Google wasn't much help either.  I'll run it for a while and see what happens.  I'll probably add a lock file to the script to be sure multiple copies of the script don't get executed.

                                      --
                                      Regards,
                                      David McAnally, WD5M
                                    Your message has been successfully submitted and would be delivered to recipients shortly.