Hi,

> Am 04.07.2016 um 15:14 schrieb Nacheva, Eva <[email protected]>:
> 
> Dear Reuti, 
> 
> In more detail, we are trying to use the prolog and epilog to gather 
> information about each job and send this information to a database which is 
> connected to a web server for easier visualisation. The whole idea of the 
> project is to be able to run large data sets via pipeline and get information 
> about the progress of the job. The information we need from the prolog and 
> epilog is as follows: 
> - in prolog - the name of the job, the start date and time, on which node, 
> etc. Just basic information about the job 
> - in epilog - if the job is successfully finished or has failed and if 
> failed, on which step; how much memory and CPU were used by the job, what is 
> the end date and time and how long did the job took to be finished.
> However, as we are not admin of the cluster, this information has to be 
> extracted without modifying the queue. Also, we cannot create a new queue for 
> us only. Moreover, as we are currently using dependencies between the jobs to 
> run the large data sets, we want to avoid adding new dependencies between 
> every job.
> We want a specific prolog/epilog for a sub set of job running a queue which 
> is already defined by the admin.

Did you have a look at the `qacct` command? Could this be a starting point to 
gather some information? `qacct -j <wc_job_id>` delivers some of the details 
which you mentioned above.

Specifying an individual prolog/epilog without changing any settings of the 
queue or the cluster is not possible.

-- Reuti


> Thank you very much in advance.
> 
> Regards, 
> Eva Nacheva
> From: Reuti <[email protected]>
> Sent: Monday, July 4, 2016 5:08:07 AM
> To: Nacheva, Eva
> Cc: [email protected]
> Subject: Re: [gridengine users] Prolog and Epilog options
>  
> Hi,
> 
> Am 04.07.2016 um 00:43 schrieb Nacheva, Eva:
> 
> > I would like to ask if there is a prolog and epilog option associated with 
> > a job in sge. I read some information from 2008 that stated this option is 
> > associated with a queue only and not a single job but this is going to 
> > change for future releases. However, as this information was written in 
> > 2008, I have been wondering if this has already been changed and updated.
> 
> Without further knowledge what you refer to with "associated" I can only give 
> a vague answer: both is true.
> 
> - A prolog/epilog is attached to a queue (or on a global level in SGE's 
> configuration).
> 
> - This prolog/epilog will be executed before resp. after each individual job.
> 
> What behavior do you expect in detail and what behavior do you think is 
> implemented right now? Do you want to specify a different prolog/epilog for 
> each job during submission time?
> 
> -- Reuti


_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to