Waiting to be Processed (Sleeping) - looking for sleeping group RRS feed

  • Question

  • Hello,

    I'm experiencing a weird behavior with the queue system, and can't figure out what the problem is. We have a farm environment of several nodes, but even when I only activate Project Server on one node, I get the same problem. On each action that uses the queue system, I get the following job state : first Waiting to be Processed, and after 2 seconds, it becomes Waiting to be Processed (Sleeping). However, when looking at the ULS logs, I can see this message passing by :

    PSI: [QUEUE] ProjectQ: Group is sleeping: 99924b6c-4b55-4d47-b132-289c5f965f00 type = ProjectUpdate

    The fact is that nothing should be sleeping at this time. All project files are published fine, SharePoint sites are working fine etc. Only each time I have to pass through the queue system, it takes for ages. I already tried the report DB resync, but that also took a long time, and finally, it succeeded.

    We have several instances running on the Project Server service application, and it all worked fine up to a few days ago. I already patched it to the FEB 2012 CU, but without any difference. I have the same behavior on any instance I try to do something with.

    How can I look for jobs that SharePoint considers to be in a sleeping mode ? And how do I get rid of them ?

    Anyone an idea ? I'm going nuts here.

    Thursday, March 8, 2012 1:10 PM

All replies

  • It seems that I have this problem only on the project queue, and not on the timesheet queue.
    Thursday, March 8, 2012 1:39 PM
  • Hi there,

    Did you try Restarting queue service? Check time on SQL server and all servers in the farm.

    Hrushikesh Deshpande – Senior EPM Infrastructure Consultant,

    Thursday, March 8, 2012 5:25 PM
  • Yes, checked it. Everything is fully in sync with my DC.

    Restarted the services and all, but still some job remains sleeping in some database. And I can't find which one... I guess it should finallly point to some job in the config database ?

    Thursday, March 8, 2012 5:32 PM
  • How about restarting all servers once? (including SQL)

    Hrushikesh Deshpande – Senior EPM Infrastructure Consultant,

    Thursday, March 8, 2012 5:39 PM
  • Tried that this night, and when it didn't seem to help, I deployed even the latest CU that was released this week.

    But what I am wondering is how someone can detect which job is causing the queue to go into a sleeping mode ? in his polling mecanism, he must fetch his state from some database table, no ?

    Thursday, March 8, 2012 5:41 PM
  • Yes, If you add sucess coloumn in queue view from "Job completion state" you could probably understand which jobs are processed by review jobs around sleeping jobs.

    My guess is reproting database. Try updating DB statistics by executing SP_UpdateStats command. (You can execute this on all 4 Project databases)

    If you provide additional ULS logs that could help.

    Hrushikesh Deshpande – Senior EPM Infrastructure Consultant,

    Thursday, March 8, 2012 5:45 PM