You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@helix.apache.org by "Hunter L (JIRA)" <ji...@apache.org> on 2018/07/24 21:12:00 UTC
[jira] [Created] (HELIX-746) [TASK] Fix removeJob so its behavior
is more consistent with removeWorkflow
Hunter L created HELIX-746:
------------------------------
Summary: [TASK] Fix removeJob so its behavior is more consistent with removeWorkflow
Key: HELIX-746
URL: https://issues.apache.org/jira/browse/HELIX-746
Project: Apache Helix
Issue Type: Improvement
Reporter: Hunter L
Change the behavior of removeJob() so that it's more consistent with removeWorkflow(). This RB addresses the scenario: suppose config deletion failed (so config still exists) and context deletion succeeded. Then the Controller has no way of knowing whether this job has ever been scheduled or it was meant to be deleted (although failed due to partial deletion). Returning as soon as config deletion fails can prevent this scenario.
Changelist:
1. Make removeJob() return early as soon as a ZNode write failure is detected
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)