linghengzhu opened a new issue #5940:
URL: https://github.com/apache/dolphinscheduler/issues/5940


   **Describe the bug**
   This bug is also reported in 
[https://github.com/apache/dolphinscheduler/issues/5528](url) which is not 
resolved yet though already closed. I may have found the real cause of this 
bug, and give detailed reproduce steps as followed.
   The bug is described as below.
   After clicking run button and choose default worker group, tasks can not be 
dispatched forever.
   Master node print error log which says fail to execute : %s due to no 
suitable worker.
   
   **To Reproduce**
   To reproduce, you can add Thread.sleep(60000) after line "load();" in method 
ServerNodeManager.afterPropertiesSet() to enlarge the timespan and increase the 
probability that this bug happens. And start master node first till it falls 
asleep, then start worker node.
   
   **Which version of Dolphin Scheduler:**
    -[1.3.6-release]
   
   **Requirement or improvement**
   From the reproduce steps, you can see the cause of this bug. It's because 
the map workerGroupNodes fail to sync with zookeeper. And that is because the 
listener registered too late. So maybe the solution is to switch the execute 
order of load() and listener registration.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to