github-actions[bot] commented on issue #16454:
URL: 
https://github.com/apache/dolphinscheduler/issues/16454#issuecomment-2287564482

   ### 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
   
   The current deployment mode is 1 master and 3 slaves.
   1. Configure the st task through ds
   2. Stop all three work-servers
   3. Start 3 worker-servers in sequence
   The following problem occurs
   1. The three worker-servers were down and did not kill the st task. The main 
reason for this is that ds is only responsible for submitting tasks to st. The 
actual task execution is run by the st server. However, when the work-server is 
started again, the previous task will be found. If the task stops unexpectedly, 
a new task will be restarted. At this time, the original st task will be 
doubled, and then the CPU and memory will be full. There will be two identical 
tasks in ds, one is running and the other is in status. It requires fault 
tolerance
   
   ### What you expected to happen
   
   ds's task monitoring for st is not complete yet. When launching a new st 
task, it did not go to stserver to check the actual running status of the task.
   
   ### How to reproduce
   
   1. Configure the st task through ds
   2. Stop all three work-servers
   3. Start 3 worker-servers in sequence
   
   ### Anything else
   
   _No response_
   
   ### Version
   
   3.2.x
   
   ### Are you willing to submit PR?
   
   - [ ] 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