[jira] [Commented] (KAFKA-10553) Track handling of topic deletion during reassignment

["Jason Gustafson (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KAFKA-10553?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17204285#comment-17204285 ] 

Jason Gustafson commented on KAFKA-10553:
-----------------------------------------

Thanks for filing the issue. I agree this is an important issue to fix. Use cases such as in testing which call for deletion and recreation of the same topic are difficult to support at the moment. I believe one of the original motivations for this behavior was due to the fact that acknowledgements were required from all replicas before deletion could complete. But if we were in a situation where the replica was permanently down, then we would need to reassign it first before deletion could proceed. So rather than trying to decide in which situations deletion could still be done while a reassignment is in progress, the code just forbids it.

> Track handling of topic deletion during reassignment
> ----------------------------------------------------
>
>                 Key: KAFKA-10553
>                 URL: https://issues.apache.org/jira/browse/KAFKA-10553
>             Project: Kafka
>          Issue Type: Sub-task
>            Reporter: Justine Olshan
>            Priority: Major
>
> Currently deletion is blocked during partition reassignment, but KIP-516 and topic identifiers can allow deletions to proceed without blocking. 
> In addition we should be able to avoid blocking deletion when a replica is down.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)