[
https://issues.apache.org/jira/browse/STORM-3586?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17044544#comment-17044544
]
Aaron Gresch commented on STORM-3586:
-------------------------------------
The scheduling timeout is not a good solution, as this could still happen with
lots of topologies running. The best thing to do would be to handle the
missing blob.
> Prevent deletion of topology blobs during initial scheduling
> ------------------------------------------------------------
>
> Key: STORM-3586
> URL: https://issues.apache.org/jira/browse/STORM-3586
> Project: Apache Storm
> Issue Type: Bug
> Reporter: Aaron Gresch
> Assignee: Aaron Gresch
> Priority: Minor
>
> If a cluster has a long scheduling period compared to the blob cleanup
> interval, it's possible to have a user submit a topology, kill it, and have
> the blob removed before supervisors get the scheduling assignment. If this
> happens, the supervisors will get FNF exceptions looking for the blob.
>
> A simple fix is to just have the blob cleanup interval be larger than the
> scheduling timeout.
>
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
