BLUF: I want to change the execute directory for multiple nodes in our cluster, but the config changes I've used do not seem to be working.
Background:
- Drive C on some of the nodes does not have a lot of space, executable is not checkpointing, but is creating several GB of intermediate files, and sending several 10^2MB back to submitting node. So I need to ensure I don't crush the system drive on these nodes.
What I've done so far:
- Searched mail list archive and got no love--bad search skills?
- Execute dir line in config file was: EXECUTE = $(LOCAL_DIR)/execute, and is now commented out
- Execute dir line is now: EXECUTE = G:\CondorExecute. I used the backslash following the convention in the config file defining where the condor dir is on the system disk. I've also tried the unix-style forward slash, and did perceive an effect either.
- Drive G in this case is an attached drive, directory has write and modify perms for users as well as SYSTEM
-issued a condor_reconfig command after change
- jobs still being executed in c:\condor\execute instead of the drive g:\CondorExecute
Anyone able to teach me where I am going wrong?
mlanham@xxxxxxxxxx
michael.lanham@xxxxxxxxxxx
5117 Wean Hall, CMU,5000 Forbes Avenue, Pittsburgh, PA 15213
412-268-4681 (w) 719-201-5520 (c)
_______________________________________________
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:
https://lists.cs.wisc.edu/archive/condor-users/