1 Reply Latest reply on May 31, 2016 11:13 AM by Jeff Strauss

    Background jobs concurrency

    Chaitanya Somanchi



      We have created a query for calculating the concurrency of the background jobs. Interestingly we are seeing that during some period the concurrency for extract refresh is above 50 -100. Which is ridiculous because at a moment of time there should be either one or two extract refreshes.

      Below is the query which i'm using.


      SELECT h.completed_at, count(id) as count from _background_tasks h,

      (SELECT distinct completed_at as tstamp from _background_tasks WHERE started_at > '4/18/2016' AND completed_at < '12/18/2016') t

      WHERE started_at > '4/18/2016' AND completed_at < '12/18/2016' AND

      t.tstamp > h.started_at AND t.tstamp < h.completed_at

      GROUP BY h.completed_at


      Please look into it and help me out in finding the exact concurrent jobs

        • 1. Re: Background jobs concurrency
          Jeff Strauss

          I'm not really sure what you're trying to accomplish, because the degree of # of backgrounders running at once is set by your server configuration of how many backgrounders are established and then whether you have established your schedules to run tasks in parallel or serial.


          But in any case, you may need to add backgrounder_id or site_id to your join in order to get proper results.