[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] restricting the number of jobs



I have a similar issue (no I/O intensive but a reasonably large number of jobs).

I set up a dagman job to control it since you can provide a command line option to condor_submit_dag that allows you to specific "maxidle" and "maxjobs". I think (from what I read), "maxjobs" only applies to each separate "job" inside the dagman and not the jobs inside the individual submit script (such as you have here) but I believe that the documentation says that "maxidle" does apply to each individual job submitted even though it is within a single submit script (such as you have).

I haven't tried it though. I just created a dagman job that would have all of the jobs separated out in the dagman submit script because I wanted to vary the input/output file names and initial directory per job as well as control how many jobs were submitted at any one time.

Maybe you can try that.

Kim

------------------------------------------------------------------------------
Kim Dillman
Research Programmer – Rosen Center for Advanced Computing
Purdue TeraGrid Campus Champion
YONG 956
Phone: 765-494-5446
Email: kadillma@xxxxxxxxxx

-----Original Message-----
From: condor-users-bounces@xxxxxxxxxxx [mailto:condor-users-bounces@xxxxxxxxxxx] On Behalf Of Mag Gam
Sent: Saturday, October 10, 2009 2:06 PM
To: Condor-Users Mail List
Subject: Re: [Condor-users] restricting the number of jobs

how can I let the user control it?

WE have some users whose jobs are very I/O intensive. They want to run
only 10 at time.



On Sat, Oct 10, 2009 at 2:55 PM, dawnsong <dawnsong.tsinghua@xxxxxxxxx> wrote:
> set
> MAX_JOBS_RUNNING = 10
> in Condor global configuration file.
>
> 2009/10/10 Mag Gam <magawake@xxxxxxxxx>
>>
>> Is it possible to restrict the number of jobs to run?
>>
>> For example?
>>
>> I have something like this:
>>
>> Universe       = vanilla
>> Executable     = hello_world.sh
>>
>> input   = /dev/null
>> output  = hello.out
>> error   = hello.error
>>
>> Queue 5000
>>
>>
>> What is I want only 10 to run at a time?
>>
>> Is that possible to do?
>> _______________________________________________
>> Condor-users mailing list
>> To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
>> subject: Unsubscribe
>> You can also unsubscribe by visiting
>> https://lists.cs.wisc.edu/mailman/listinfo/condor-users
>>
>> The archives can be found at:
>> https://lists.cs.wisc.edu/archive/condor-users/
>
>
> _______________________________________________
> Condor-users mailing list
> To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
> subject: Unsubscribe
> You can also unsubscribe by visiting
> https://lists.cs.wisc.edu/mailman/listinfo/condor-users
>
> The archives can be found at:
> https://lists.cs.wisc.edu/archive/condor-users/
>
>
_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users

The archives can be found at: 
https://lists.cs.wisc.edu/archive/condor-users/