none
Jobs stuck in queued state RRS feed

  • Question

  • Hi,

    Same problem is described on MS forum. http://social.microsoft.com/Forums/en-US/6392c7a2-5fc2-4f21-801f-82a25ad31253/slow-dispatching-or-ui-updates-for-large-jobs?forum=windowshpcsched. But we can not find a solution of this issue and possible reasons in this topic.

    We use hpc 2008 r2 server. Cluster consist of 6 compute nodes (each one has 12 cores and the same memory size) and 1 head node.

    We run a large Job with over 2000 tasks. All the tasks have different names and each one requests for one core. First 1800 of them were executed quite fast and without interruptions. But then the filter by "running" state doesn't show any running tasks but at the same time there are about 200 tasks in "queued" state. This can continue over 15 minutes. Next tasks are started and it repeats again.

    Please advise what can we do with the issue to resolve it.

    Friday, October 18, 2013 2:37 PM