It sounds like you have other tasks already executing on your 4 backgrounders, and therefore "this" refresh task doesn't start immediately as scheduled. This is pretty much expected if you have too much work going through too few backgrounders.
Even if you DO raise priority on this task, it simply means it'll "go first" when a backgrounder opens up. However it won't do anything to ensure the task starts executing immediately, which I think is probably what your customer wants. A backgrounder must be free in order for work to start immediately. That means you need to do less refreshing work elsewhere, or have more backgrounders (and therefore more hardware, more Tableau, etc.)
What should you do?
- Understand that only so many extract refresh processes can execute at the same time (in your case, 4). Make sure your customer understands that a backgrounder is a very limited resource and they will "use them up" if they use refresh schedules that execute too often. Question them closely about how often each of their refresh schedules REALLY needs to execute keeping in mind the upper limit it seems they're already hitting. Be firm
- Consider creating two schedules - One with a higher priority which runs tasks in parallel. Only refresh jobs that execute "quickly" should be assigned to this schedule. Create a second schedule with a lower priority - set it to run tasks serially. Assign "slow" refresh jobs to it. This approach will prevent 4 "slow" refresh tasks from totally soaking up all your backgrounders and preventing other tasks which could execute quickly from even starting.
Thanks! It is very helpful for me.
I am new to Tableau server if I suggest to increase backgrounders is it helpful.
Tableau creates any Error loges for this kid of Issues and how can I monitor how many Extracts are running in same time.
The maximum number of backgrounder you should run on any machine is (# of cores / 2) - So, if you have an 8 core machine, 4 backgrounders.