[torqueusers] Re: pbs mom_logs. "no active process found"
garrick at usc.edu
Thu Dec 11 12:05:44 MST 2008
On Thu, Dec 11, 2008 at 12:16:03PM -0600, Rahul Nabar alleged:
> >Do you have this set in your PBS server ?
> >set server mom_job_sync = True
> >It's meant to get the pbs_mom to remove jobs on
> >compute nodes when they've already gone from the
> Thanks Chris! I issued a "print server" on my master-node. No, I
> could not find the setting you mention. I can give that a shot.
Settings aren't visible unless they are "set". See the pbs_server_attributes
manpage for the full list.
> But here's a risk though: Currently we can momentarily lose
> connectivity with the pbs_server and the compute nodes will continue
> to do their jobs. Will this new setting have the side effect of
> flushing all jobs from the compute nodes if they cannot talk to the
> master-node / the pbs_server? That is the interpretation of "already
> gone from the server"?
No. If set, pbs_server will tell pbs_mom to remove jobs when they don't exist
on the server.
Garrick Staples, GNU/Linux HPCC SysAdmin
University of Southern California
See the Dishonor Roll at http://www.californiansagainsthate.com/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20081211/d4e28748/attachment.bin
More information about the torqueusers