I am currently trying to setup a Retry time cycle (R100/PT3S) in an async Service Task as shown in the image below:
As per my understanding of the documentation, the expected behaviour for R100/PT3S should be:
- 100 max retry attempts
- 3 second timeout after each attempt
However, I notice that the engine does not honour the 3 seconds timeout, and instead, performs a timeout for atleast 5 seconds (as an additional weird behaviour, the engine seems to be performing timeouts that vary between 5-7 seconds).
- Is there a way to set a timeout of less than 5 seconds?
- Maybe provide the retry time cycle in a different format somehow? (e.g. in milliseconds) to apply the desired behaviour.
Hello my friend! Welcome to forum again!
I’m not sure, but I believe that this behavior could be caused by the Job Executor’s delay in picking up the task to execute, and not by the specific retry time, and you can configure this if you wish, remembering to always carry out the tests before uploading in production to check if this is actually the expected behavior
Below I am leaving Camunda’s official documentation related to Job Executor so that you can check it better.
Hope this helps.
William Robert Alves
Hello William, thanks for your answer
After reading the Job Executor documentation and all the properties mentioned, it still is not clear to me which Job Executor configuration property is related with the behaviour I described.
As a side note, this behaviour I described does not seem to be relevant with the load the application has. Even if only 1 process instance is running, the engine performs Retry intervals of at least 5 seconds.