Hi, several months ago when I started learning Slurm and reading through
the web pages, I made this picture to help myself understanding the *prolog
and *epilog interactions with job steps. Please see the attachment. If you
see any corrections necessary, please inform. Thank you!


Best Regards.

On Thu, Apr 13, 2017 at 3:35 AM, Chrysovalantis Paschoulas <
c.paschou...@fz-juelich.de> wrote:

> Hi Roger,
>
> instead of canceling the job normally, the user could use "scancel
> --signal=<signal_name> <jobid>" in order to send a specific signal to his
> jobscript or application.
>
> His application/jobscript now should be able to handle that signal and
> after performing the cleanup tasks it can terminate its self gracefully.
>
> The Epilog scripts are something else now. There are Prolog/Epilog scripts
> that run as root, there are Task-Prolog/Epilog script that run as normal
> user but are defined system-wide by sysadmins and there the per-srun user
> prolog/epilog scripts defined with srun's option --task-prolog and
> --task-epilog which should be short otherwise killed.
>
> Also you could have a look into slurm.conf options like: KillWait,
> WaitTime and all the Timeouts (not all of them are useful for your case
> though).
>
> Best Regards,
> Chrysovalantis Paschoulas
>
>
>
> On 12.04.2017 23:40, Roger Moye wrote:
>
> I have an unusual configuration question for the Slurm community.
>
>
>
> When a user cancels a job he sees this message:
>
> srun: Job step aborted: Waiting up to 2 seconds for job step to finish.
>
>
>
> Is there a way to lengthen this time?
>
>
>
> One of our users has constructed his jobs such that when a job is
> cancelled, it tries to perform some cleanup tasks before exiting but these
> tasks take longer than 2 seconds.
>
>
>
> Is there a way to lengthen the time a job will continue to run even after
> it has been cancelled?   I guess this would be long enough to allow an
> Epilog script to run.
>
>
>
> Thanks in advance!
>
> -Roger
>
>
>
>
>
> [image: cid:image001.png@01D22319.C7D5D540]
>
> Roger Moye
>
> HPC Engineer
>
> 713.425.6236 <(713)%20425-6236> Office
>
> 713.376.2540 <(713)%20376-2540> Mobile
>
>
>
> QUANTLAB Financial, LLC
>
> 3 Greenway Plaza
>
> Suite 200
>
> Houston, Texas 77046
>
> www.quantlab.com
>
>
>
>  -----------------------------------------------------------
> ------------------------
>
> The information in this communication and any attachment is confidential
> and intended solely for the attention and use of the named addressee(s).
> All information and opinions expressed herein are subject to change without
> notice. This communication is not to be construed as an offer to sell or
> the solicitation of an offer to buy any security. Any such offer or
> solicitation can only be made by means of the delivery of a confidential
> private offering memorandum (which should be carefully reviewed for a
> complete description of investment strategies and risks). Any reliance one
> may place on the accuracy or validity of this information is at their own
> risk. Past performance is not necessarily indicative of the future results
> of an investment. All figures are estimated and unaudited unless otherwise
> noted. If you are not the intended recipient, or a person responsible for
> delivering this to the intended recipient, you are not authorized to and
> must not disclose, copy, distribute, or retain this message or any part of
> it. In this case, please notify the sender immediately at 713-333-5440
> <(713)%20333-5440>
>
>
>
>
> ------------------------------------------------------------
> ------------------------------------
> ------------------------------------------------------------
> ------------------------------------
> Forschungszentrum Juelich GmbH
> 52425 Juelich
> Sitz der Gesellschaft: Juelich
> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
> Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
> Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
> Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
> Prof. Dr. Sebastian M. Schmidt
> ------------------------------------------------------------
> ------------------------------------
> ------------------------------------------------------------
> ------------------------------------
>
>

Reply via email to