[Condor-users] Condor on XP stops talking to pool manager

[Andrew Mellanby <A.Mellanby@xxxxxxxxxxxxx>]

Hi,
I've installed 28 WindowsXP machines with condor 6.6.10.
After the 1st day of running, 11 have stopped showing up on the condor_status 
list.

The MasterLog on some of the affected machines ends with this:
> 8/30 18:17:33 Child 2424 died, but not a daemon -- Ignored

When I check the processes running on the afflicted system, they do appear to 
be active, but I can't contact them from the pool manager

> condor_restart -direct 130.195.109.42
> 9/1 10:04:19 TOOL_TIMEOUT_MULTIPLIER is undefined, using default value of 0
> Can't find address for master 130.195.109.42

Is there anything else I can do to get these machines back online without 
actually rebooting them ?


thanks

Mel.



8/30 17:17:32 ******************************************************
8/30 17:17:32 ** Condor (CONDOR_MASTER) STARTING UP
8/30 17:17:32 ** C:\Condor\bin\condor_master.exe
8/30 17:17:32 ** $CondorVersion: 6.6.10 Jun 22 2005 $
8/30 17:17:32 ** $CondorPlatform: INTEL-WINNT50 $
8/30 17:17:32 ** PID = 2196
8/30 17:17:32 ******************************************************
8/30 17:17:32 Using config file: C:\Condor\condor_config
8/30 17:17:32 Using local config files: C:\Condor/condor_config.local
8/30 17:17:32 DaemonCore: Command Socket at <130.195.109.42:3022>
8/30 17:17:32 Started DaemonCore process "C:\Condor/bin/condor_startd.exe", 
pid and pgroup = 916
8/30 18:17:32 Preen pid is 2424
8/30 18:17:33 DaemonCore: Command received via UDP from host 
<130.195.109.42:3182>
8/30 18:17:33 DaemonCore: received command 60001 (DC_PROCESSEXIT), calling 
handler (HandleProcessExitCommand())
8/30 18:17:33 Child 2424 died, but not a daemon -- Ignored