Re: [HTCondor-users] How to make condor auto-reschedule my jobs executing on nodes which are failed because of hardware?

[钱晓明 <kyleqian@xxxxxxxxx>]

I think I can use a script to hold jobs executing on this machine before shutting down and release them later by schedd.

Is there any configuration for condor daemons to do this?

2013/9/9 钱晓明 <kyleqian@xxxxxxxxx>

by the way, signal 9 is caused by shutting down the compute node.

2013/9/9 钱晓明 <kyleqian@xxxxxxxxx>

Hi, I found condor would not reschedule my jobs executing on nodes which were failed because of hardware or power. I think there is a way to tell condor do it. Can anyone point it to me?  Or I must monitor the log file and do it by myself?

I also found that if a node shutdown, jobs executing on it would be terminated abnormally by signal 9.

 

I hope that the solution can apply to cluster job. Thanks very much!