[torqueusers] Attn: Accelrys Materials Studio users

Chris Samuel csamuel at vpac.org
Wed Feb 8 15:28:18 MST 2006


On Friday 20 January 2006 12:07, Steven A. DuChene wrote:

> and because of issues with Materials Studio reporting a job as dead while
> it was still running on a Linux cluster.

Ah, this would be the bug that if qstat failed for any reason at all (such as 
a timeout talking to the pbs_server) then *they* assumed that the job had 
died, whereas Torque was still quite happily tracking it.

Strangely I managed to fix this alleged bug in OpenPBS by patching their code 
to do the right thing (only report a job as dead if qstat returned the code 
for no such job ID).

They told me my patches to fix this would be incorporated into version 4.0.

cheers,
Chris
-- 
 Christopher Samuel - (03)9925 4751 - VPAC Deputy Systems Manager
 Victorian Partnership for Advanced Computing http://www.vpac.org/
 Bldg 91, 110 Victoria Street, Carlton South, VIC 3053, Australia

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20060209/3268e483/attachment.bin


More information about the torqueusers mailing list