Killing a job in SQL should free up thread in Service

I've noticed a few times that sometimes when you kill a job in SQL, it still occupies one of the threads available from the service. There is no visibility into this. So if you kill couple of jobs throughout the week, you could see all the services as up and running, but the job queue shows jobs are queued up and not processing (and with no task currently processing). Is it possible to have the SQL failure cascade up and release the thread on the service? 

  • Eric Weldon
  • Apr 26 2018
  • Future consideration
Stewardship Tier (SST) Framework
  • Attach files
  • Tim Empson commented
    11 Sep, 2018 09:34am

    This is particularly a problem with services.  We have a service that ocasionally goes rouge (Send Workflows).  This happens when a collect connection fails and we have many many e-mails to send.  The service doesn't finish before the next starts and quickly they fill the queue.
    Freeing up the slots on the queue so other jobs can restart is problematical.

    Attachments Open full size
  • Admin
    Jon Green commented
    18 May, 2018 10:08am

    Thanks for logging the idea Eric, we will investigate whether this would be possible in the DSP framework so that this improvement can be implemented.

    Attachments Open full size

Related ideas