Jump to content
  • Announcements

    • admin

      PBS Forum Has Closed   06/12/17

      The PBS Works Support Forum is no longer active.  For PBS community-oriented questions and support, please join the discussion at http://community.pbspro.org.  Any new security advisories related to commercially-licensed products will be posted in the PBS User Area (https://secure.altair.com/UserArea/). 
Sign in to follow this  
Rodrigo

Job would conflict with starving job

Recommended Posts

Hello, anyone know why this problem? i don't know why the job is suspended or enqueued.


# tracejob 5558.cluster

Job: 5558.cluster

02/11/2012 10:23:01 S Job Queued at request of user1@cluster, owner = user1@cluster, job name = test2, queue = medium
02/11/2012 10:23:01 S Job Modified at request of Scheduler@cluster
02/11/2012 10:23:01 S Job Modified at request of Scheduler@cluster
02/11/2012 10:23:01 S Job Modified at request of root@cluster
02/11/2012 10:23:01 L Job would conflict with starving job
02/11/2012 10:23:01 S enqueuing into default, state 1 hop 1
02/11/2012 10:23:01 S dequeuing from default, state 1
02/11/2012 10:23:01 S enqueuing into medium, state 1 hop 1
02/11/2012 10:29:57 S Job Modified at request of root@cluster
02/11/2012 10:29:57 L Job is starving
02/11/2012 10:29:57 L Job would conflict with starving job

just means I have more than a starving job?

Att.

Share this post


Link to post
Share on other sites

Hi Rodrigo,

It is hard to give you advice on what is going on because I don't know what you are trying to accomplish.

Looks like the job is now starving, but because I don't know what your sched_config looks like or what other jobs are in the system, it is tough to describe what you are seeing.

If you look at the sched_config, you will see parameters for starving jobs... when this is enabled, then the scheduler will begin to drain the system to make sure the 'starving job' runs. Now, in order for starving jobs to work well, then the jobs in the system should be requesting wall time. This way the scheduler can simulate the future/calendar and forecast what is going to happen to jobs (e.g., preemption).

The enqueueing references means the job was put into the queue. Looks like you may have routing queues? job went into default and then bounced in medium?

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this  

×