[gridengine users] qstat -s p +qstat -r <> qstat -s rp
suhovecky at nd.edu
Tue May 17 14:38:02 UTC 2011
How many jobs does your scheduler
allow to run at once per user?
qconf -ssconf (look for maxujobs)
I've had users schedule task arrays with 40K tasks,
but we set maxujobs to between 100-200, and that's
as many as will run at once.
From: users-bounces at gridengine.org [users-bounces at gridengine.org] On Behalf Of Hung-ShengTsao (Lao Tsao) Ph.D. [laotsao at gmail.com]
Sent: Tuesday, May 17, 2011 10:21 AM
To: William Hay
Cc: users at gridengine.org
Subject: Re: [gridengine users] qstat -s p +qstat -r <> qstat -s rp
how many total slots in your cluster for 5000 tasks?
On 5/17/2011 10:19 AM, William Hay wrote:
> On 17 May 2011 15:06, Hung-ShengTsao (Lao Tsao) Ph.D.<laotsao at gmail.com> wrote:
>> I just check in a test cluster
>> qsub -t 1-10 simple.sh
>> qstat -s p -u<uid> does show qw stat with the ja-task ID changing
>> qstat -s r -u uid show running job
>> which version of ge your are running
>> I use ge6.2u5 in r5.4
> 6.2u3 (we should upgrade). We also have array jobs from other jobs
> where both running and queued tasks show up.
>> On 5/17/2011 9:51 AM, William Hay wrote:
>>> One of our users has submitted an array job of about 5000 tasks.
>>> Some of these tasks are running. Some are not. If I run qstat -s rp
>>> -u<username> then I see tasks in both states "r" or state "qw".
>>> Running qstat -s r -u<username> shows the running tasks. However
>>> running qstat -s p -u<username> does not show the tasks in state qw.
>>> These waiting tasks had previously been in an error state caused by a
>>> problem with our cluster file system but I cleared the error once we
>>> had fixed
>>> the problem. Any idea what is going on and whether these tasks will
>>> eventually run.
>>> users mailing list
>>> users at gridengine.org
More information about the users