[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Condor-users] Condor Error No locks available
- Date: Tue, 13 Jun 2006 12:50:48 -0500 (CDT)
- From: "R. Kent Wenger" <wenger@xxxxxxxxxxx>
- Subject: Re: [Condor-users] Condor Error No locks available
On Tue, 13 Jun 2006, lohit wrote:
> This worked, Thanks!
> I moved log files for individual jobs to local disk. But I am not clear
> about the concept, becuase the individual log files had different filenames.
> and there were no common log files.
I'm not 100% clear on what you're asking here, but I'll do some
explaining, and you can ask for clarification if I'm not answering
the right question...
Anyhow, each node job has a log file defined in its condor_submit file.
That log file is written by condor_submit, the schedd, shadow, etc.
The log file is also *read* by DAGMan itself -- that's how DAGMan monitors
the progress of the node jobs.
Both the log reading and writing code gets a lock on the relevant log
file, basically so that partial events are not written or read. In
general, we have seen that the locking doesn't work right when the log
files are on NFS.