HelloÂChristoph,
May be not directly answering your question but thought this may provide some help:
Recently I came across the following command which I found very useful to get the history of jobs ran on the executor node. You need to fire this command on the executor node. It's very useful to see the jobs ran on the node submitted from different schedulers during the time of issue for troubleshooting purposes. It covers history not current runs.Â
condor_history -file `condor_config_val LOG`/startd_history -limit 2 -af remotehost globaljobid
We have clusters consisting of 400+ nodes. We do capture condor_who at intervals of 1 minute and it doesn't seem to be causing any issue for us.Â
Thanks & Regards,
Vikrant Aggarwal