Thread Pool (Asynchronous Runner)

Owned by Jonathan Muse
Last updated: Nov 19, 2024
3 min read

 

The Thread Pool Configuration (formerly Asynchronous Runner) is located in Power Apps Configuration page in the Managed Apps page of the Jira Administration.

To find it, click the General menu and open the Thread pool sub-menu.

Thread Pool Configuration

Various tasks require running SIL™ programs in the background. These tasks are run in a special thread pool where each script is run as a separate process. Thread pool configuration is easy, however you must understand the implications of each parameter since they may affect the tasks that are run in the background.

Threads

The number of running threads (number of SIL™ scripts running in the same time).

  • A higher number of threads means that scripts will complete faster because more scripts can run at the same time. However, this puts more stress on the systems processor and memory.

  • A lower number of threads means that some scripts may need to wait a few milliseconds before they can be processed. In turn, this results in less stress to the systems performance.

Time to Live (TTL)

Time to live (running time for a SIL™ script). If you run a task that takes more than TTL configured, the script ends with no result.

This protects Jira from scripts that may be waiting on external systems to complete.

  • A higher Time to Live means that scripts will have ample time to complete their allotted task. However, this means that resources (processors and memory) could be tied up for longer.

  • A lower Time to Live means that resources (processors and memory) will be released faster helping ensure a high performant system.

Checkpoint Interval

Interval to clean up the expired tasks (the SIL™ scripts that exceeds the TTL configured). The pool is maintained by a watchdog thread, and the checkpoint interval determines the interval between checks.

  • A lower interval means that the watchdog will check more often and that resources can be returned to Jira faster. However, the watchdog itself does require resources to run, extremely minimal resources but resources none the less.

  • A higher interval means that the watchdog will check less often and it may take a few more minutes for resources to be returned. However, this means that the watchdog will take up less resources overall.

Keep in mind that the Time to Live setting and the Checkpoint Interval are determined in minutes and hours while most scripts can complete in a few milliseconds.

Tips

The page shows also the maximum task queue length and the current tasks that are running. If your number of threads is constantly under the the max reached queue length, consider increasing the thread pool size.

If you run reporting tasks, that may gather data from multiple sources and take a long time to complete, please increase both the number of threads (if the pool is occupied with those tasks) and the TTL so that the tasks/results will not be removed too early from the queue.


More Configuration Guides