dolphinscheduler: [Bug] [Master] resume failure processInstance state remain failure

Search before asking

  • I had searched in the issues and found no similar issues.

What happened

Workflow process instance resume from failure state, and all task run success actully, but task instance and process intance remain failure. And when you click resume again, all switch to success.

Checking task log, the task has submit and run success when resume.

Checking database table t_ds_task_instance. There two record, old one state is failure, and new one is success. Which make sense. 企业微信截图_20221224173844

Since task is success, why ds judge it faliure and show failure on ui. Digging the ui http query, we found that it query both two record and return it, because of the flag. 企业微信截图_20221224174251 企业微信截图_20221224174335 企业微信截图_20221224174406 企业微信截图_20221224174501

And I don’t know what happen, thx for help.

What you expected to happen

both task instance and process instance should change to success

How to reproduce

  1. create a workflow with task which suppose to run fail, and configure retry 3 times
  2. run workflow. task instance state should be failure and waiting to retry, and process instance stay running
  3. click stop. task instance state should be failure and process instance state should be stoped
  4. modify task make it will run success
  5. click resume
  6. task has submit and run. but when it finish, both task and process instance state change to failure
  7. click resume again
  8. both task and process change to success

Anything else

we are on v2.0.7

Version

2.0.x

Are you willing to submit PR?

  • Yes I am willing to submit a PR!

Code of Conduct

About this issue

  • Original URL
  • State: closed
  • Created 2 years ago
  • Comments: 17

Most upvoted comments