Re: [Condor-users] condor_collector starting and dying without even writing to the log

[Rob Pieké <pieke@xxxxxxxxxx>]

Aha! Our sys-admins tracked the problem down to a missing compatability 
library (for libstdc++). I still don't understand why condor_collector 
could be launched from the command-line, but our setup is back up and 
running, so I won't look a gift-horse in the mouth.



> This is for the RH9 "** $CondorVersion: 6.6.11 Mar 23 2006 $" build
> running on Fedora Core 4. Interestingly, we weren't seeing this for tje
> RH7 "** $CondorVersion: 6.6.9 Mar 10 2005 $" build running on Fedora Core 4.
>
>  From the MasterLog:
>
> 11/22 14:38:32 Started DaemonCore process
> "/usr/local/condor/sbin/condor_collector", pid and pgroup = 4480
> 11/22 14:38:32 enter Daemons::UpdateCollector
> 11/22 14:38:32 Attempting to send update via UDP to collector
> XXX.corefa.com <XXX.XXX.XXX.XXX:9618>
> 11/22 14:38:32 Can't connect to <XXX.XXX.XXX.XXX:9618>:0, errno = 111
> 11/22 14:38:32 Will keep trying for 10 seconds...
> 11/22 14:38:42 Connect failed for 10 seconds; returning FALSE
> 11/22 14:38:42 ERROR:
> SECMAN:2003:TCP connection to <XXX.XXX.XXX.XXX:9618> failed
>
> 11/22 14:38:42 Can't send UPDATE_MASTER_AD to collector XXX.corefa.com
> <XXX.XXX.XXX.XXX:9618>: Failed to send UDP update command to collector
> 11/22 14:38:42 DaemonCore: No more children processes to reap.
> 11/22 14:38:42 start recover timer (63)
> 11/22 14:38:42 Started DaemonCore process
> "/usr/local/condor/sbin/condor_schedd", pid and pgroup = 4485
> 11/22 14:38:42 enter Daemons::UpdateCollector
> 11/22 14:38:42 Attempting to send update via UDP to collector
> XXX.corefa.com <XXX.XXX.XXX.XXX:9618>
> 11/22 14:38:42 Can't connect to <XXX.XXX.XXX.XXX:9618>:0, errno = 111
> 11/22 14:38:42 Will keep trying for 10 seconds...
> 11/22 14:38:52 Connect failed for 10 seconds; returning FALSE
> 11/22 14:38:52 ERROR:
> SECMAN:2003:TCP connection to <XXX.XXX.XXX.XXX:9618> failed
>
> 11/22 14:38:52 Can't send UPDATE_MASTER_AD to collector XXX.corefa.com
> <XXX.XXX.XXX.XXX:9618>: Failed to send UDP update command to collector
> 11/22 14:38:52 The COLLECTOR (pid 4480) exited with status 127
> 11/22 14:38:52 ProcAPI::buildFamily failed: parent 4480 not found on system.
> 11/22 14:38:52 restarting /usr/local/condor/sbin/condor_collector in 265
> seconds
>
>
> This looks to be similar to:
> https://www-auth.cs.wisc.edu/lists/condor-users/2006-May/msg00289.shtml
>
> Any ideas on how I can diagnose this further? Some other observations:
>
> * I can launch condor_collector from the command-line, and THEN it
> writes to the log. But the process doesn't actually do any collecting
>
> * If I hammer 'ps' to see when Condor is launching condor_collector, the
> launched process is owned by 'root', not 'condor'
>
> _______________________________________________
> Condor-users mailing list
> To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
> subject: Unsubscribe
> You can also unsubscribe by visiting
> https://lists.cs.wisc.edu/mailman/listinfo/condor-users
>
> The archives can be found at either
> https://lists.cs.wisc.edu/archive/condor-users/
> http://www.opencondor.org/spaces/viewmailarchive.action?key=CONDOR