ziqiang-wang opened a new issue, #13797:
URL: https://github.com/apache/dolphinscheduler/issues/13797

   ### 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 workflow definition contains a task of a dependency type. Set Retry 
Times to 3 and Retry Interval to 1 minute. Enable Timeout Alarm, enable Wait 
for Dependency Completion, set Timeout to 1 minute, and select Timeout Policy 
to Timeout Failure, as shown in the following figure:
   
   
![image](https://user-images.githubusercontent.com/107013241/227882237-5c31de77-3271-4d4a-b876-799ddaf7a2cc.png)
   
   I then manually execute the workflow and select "Failure Policy" as 
"Continue" and execute the workflow as shown below:
   
   
![image](https://user-images.githubusercontent.com/107013241/227883038-483ed09e-f061-4514-8851-9a72bef44c8f.png)
   
   
   During execution, it was found that the dependent task did fail after a 1 
minute timeout, but after another 1 minute, it was not retried (3 failed 
retries have been set up above), and the workflow instance was always running. 
After manually stopping a workflow instance, the workflow instance remains in 
the ready to stop state.
   
   
   ### What you expected to happen
   
   1. A workflow has a dependency task. If the dependency task fails to time 
out during workflow execution, you can retry the workflow based on the set 
retry times and interval.
   2. If a dependent task fails to run eventually, the entire workflow can 
continue to run subsequent tasks or stop running directly according to the set 
running policy.
   
   ### How to reproduce
   
   Create a workflow definition where the first task type is a dependent task, 
which is dependent on any task that is not executed today, so that when the 
dependent task runs, it can time out and fail. Then enable the timeout alarm 
for the dependent task and wait for the dependency to complete, and set the 
timeout failure to the timeout policy.
   Followed by any other type of task, such as a shell.
   After saving the work definition and executing the workflow manually, you 
should see that the dependency failed to time out, but was not retried, and the 
corresponding workflow instance was always running. Manually stopping a 
workflow instance does not stop the workflow completely, but is always in a 
state of ready to stop.
   
   ### Anything else
   
   _No response_
   
   ### Version
   
   3.1.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