5 Replies Latest reply on Aug 10, 2017 10:59 AM by Jayesh Patel

    backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit


      Hi experts,

      I am looking for some more detail on "backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit" tabadmin options. I currently have 7200 querylimit which kills all jobs exceeding 2 hours but that is too much if too many jobs get stuck due to any issue - I am looking for some way that tableau can let other jobs kick in while it is waiting for backend data source to return data set.


      Just curious if using Extra_timeout_in_seconds will help?


      For example; if I set;

      .querylimit=1800 (30 mins)

      .extra_timeout_in_seconds=5400 (90 mins)


      What will happen? if a job exceeds 30 minutes, will tableau server let it run for another 90 minutes but release the Backgrounder for other jobs?


      Looking for some additional info before I try these changes.


      Thanks for your help everyone.






        • 1. Re: backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit
          Russell Christopher

          Hey Syed -


          Unfortunately, Backgrounders aren't really "smart" enough to do this. They just launch the job and wait for it to succeed for fail. Sorry.


          I saw your other question, and it sounds like you may simply need more Tableau to handle the number of extracts you're generating.

          2 of 2 people found this helpful
          • 2. Re: backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit
            SYED REHMAN

            I am really trying to understand what is Real difference between .querylimit and .extra_timeout_in_seconds.


            What will be behavioral difference in following 2 scenarios.


            Scenario 1:

            backgrounder.querylimit = 7200

            backgrounder.extra_timeout_in_seconds = "" (disabled)


            Scenario 2:

            backgrounder.querylimit = 1800

            backgrounder.extra_timeout_in_seconds= 5400



            Syed Rehman

            Visit us at http://go/sea/  or check out http://go/bi-reporting   http://go/tableau

            • 3. Re: backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit
              Justin Mangue

              Hi Syed,


              If you only want to grant extra job completion time to extract and subscription tasks, then backgrounder.querylimit is the way to go.  From the point of view of job timeout for those job types, scenario 1 & 2 are effectively the same -- after 7200 seconds, the job will be cancelled and marked as failed.


              However, scenario #2 would also change the job timeout for all other job types (syncing AD groups, indexing content, etc).  I wouldn't recommend going this route unless you are having problems with those other job types completing in time.


              As Russell mentioned, there isn't really a concept of Tableau letting a job run but "releasing" backgrounder.  If you are having problems with jobs waiting a long time to get picked up by a backgrounder, then you will probably either need to assign more backgrounders, lessen the amount of work being done, or balance your job schedules so that the load is more spread out.

              1 of 1 people found this helpful
              • 4. Re: backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit
                SYED REHMAN

                Thanks Justin... infact  .Querylimit is set/defaulted to 7200 (2 hrs) and I recently found that .extra_timeout_in_seconds is also set/default to 1800 - which was causing our extract to occupy even more then 2 hrs (more then 7200 seconds) which was puzzling for me - I infact want to restrict the max allowed time to less then 2 hrs. I know I can do it by reducing Querylimit but want to understand real implication of using extra_timeout... to make correct decision.


                Should I change .Querylimit to 5400 and.Extra_timeout... to 1800 to allow 2hrs

                or 7200, and 0 respectively will also do same job?


                couldn't find real deeper understanding of later one.

                • 5. Re: backgrounder.extra_timeout_in_seconds vs backgrounder.querylimit
                  Jayesh Patel

                  I think either you can increases the # of backgrounder or disable backgrounder.extra_timeout_in_seconds so that you notify when extract fails or times out. We are having the same issue and exploring option.