ruanwenjun opened a new issue, #10842:
URL: https://github.com/apache/dolphinscheduler/issues/10842

   ### Search before asking
   
   - [X] I had searched in the 
[issues](https://github.com/apache/dolphinscheduler/issues?q=is%3Aissue) and 
found no similar issues.
   
   
   ### What happened
   
   I find some taskInstance may not be failovered when the worker down, and the 
taskInstance's status is DISPATCH.
   
   When we do worker failover, we need to find out which taskInstances belong 
to the need failover worker, we use taskInstance's host to do this judge, if 
the taskInstance's host == needFailoveredWorker, then we may failover this 
task(We have extra judgement).
   
   <img width="351" alt="image" 
src="https://user-images.githubusercontent.com/22415594/177907335-1a5d0eb7-930d-4a98-b7aa-7de36536b740.png";>
   
   The problem is that when we success dispatch the taskInstance to a worker, 
the taskIntance's host may still null, if then the worker down, this 
taskInstance may not be failovered.
   <img width="676" alt="image" 
src="https://user-images.githubusercontent.com/22415594/177907775-d9e8b282-fa71-4beb-b3cb-7952e0e6f0ca.png";>
   
   
   
   
   ### What you expected to happen
   
   The worker failover can work will.
   
   ### How to reproduce
   
   <img width="695" alt="image" 
src="https://user-images.githubusercontent.com/22415594/177907823-72421a0f-54d3-4ff9-bf1a-ef40d830e94f.png";>
   
   Start some master and worker, and submit a lot of processInstance, then kill 
one worker. 
   
   ### Anything else
   
   _No response_
   
   ### Version
   
   dev
   
   ### Are you willing to submit PR?
   
   - [X] Yes I am willing to submit a PR!
   
   ### Code of Conduct
   
   - [X] I agree to follow this project's [Code of 
Conduct](https://www.apache.org/foundation/policies/conduct)
   


-- 
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