[gold-users] statistics (gstatement/glsjob) missing StartTime and EndTime
Scott Jackson
scottmo at adaptivecomputing.com
Mon Apr 30 14:22:48 MDT 2012
Eva,
I am pretty sure the stuff that is sent from Maui to Gold is hard-coded and
could not be changed without modifying the source code.
Your best alternative may be to use the hidden CreationTime and
ModificationTime fields. If you are doing charging, then it is likely that
the CreationTime will match up very closely with the StartTime and the
ModificationTime will match up very closely with the end time of the job.
You can do queries against this like:
glsjob --show "Id,JobId,...,ModificationTime,CreationTime"
The only other alternative I see would be to disable charging in Maui and
write your own reservations in the prolog and charges in the epilog where
you can control what you push into greserve and gcharge.
Thanks,
Scott
On Mon, Apr 30, 2012 at 1:30 PM, Eva Hocks <hocks at sdsc.edu> wrote:
>
> Hi,
>
> since using maui and gold the jobs get charged but there is no StartTime
> and
> EndTime for each job whih makes it impossible to create proper statistics
>
>
> with moab/gold:
>
> Id JobId User Project Machine Queue QualityOfService Stage
> Charge Processors Nodes WallDuration StartTime EndTime
> Description
> ------ ------ --------- ---------- ------- ----- ---------------- -------
> ------ ---------- ----- ------------ -------------------
> ------------------- -----------
> 2092 1580 xxxxx xxxxx Triton Charge
> 0 1 1 1 2009-09-22 14:50:52 2009-09-22 14:50:53
>
>
>
> with maui/gold:
>
> Id JobId User Project Machine Queue QualityOfService Stage
> Charge Processors Nodes WallDuration StartTime EndTime
> Description
> ------ ------ --------- ---------- ------- ----- ---------------- -------
> ------ ---------- ----- ------------ -------------------
> ------------------- -----------
> 981744 939245 wpfeiffer stsi-group Triton pdaf DEFAULT Charge
> 12 1 44365
>
>
>
>
> Any idea how to configure maui/gold to report the StartTime and EndTime
> per job?
>
>
> Thanks
> Eva
>
> _______________________________________________
> gold-users mailing list
> gold-users at supercluster.org
> http://www.supercluster.org/mailman/listinfo/gold-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.supercluster.org/pipermail/gold-users/attachments/20120430/c50bed8b/attachment-0001.html
More information about the gold-users
mailing list