[
https://issues.apache.org/jira/browse/IMPALA-5468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17253053#comment-17253053
]
Tim Armstrong commented on IMPALA-5468:
---------------------------------------
There have been many catalog optimisations but IMPALA-8937 would be required to
fully addrses this.
> REFRESH and a very large table run after catalog restart cannot be cancelled
> ----------------------------------------------------------------------------
>
> Key: IMPALA-5468
> URL: https://issues.apache.org/jira/browse/IMPALA-5468
> Project: IMPALA
> Issue Type: Bug
> Affects Versions: Impala 2.8.0
> Reporter: Mala Chikka Kempanna
> Priority: Major
>
> It's not possible to cancel a REFRESH on a very large table which is run
> immediately after catalog restart.
> This can be reproduced by creating a very large table in terms of metadata
> like say for eg. 4 million files with default 128 MB block size and each file
> of size 500MB.
> Trying to cancel refresh on this table that is run immediately after catalog
> restart does not succeed and the only way to stop this refresh is to restart
> the specific coordinator.
> Please make it possible to cancel long running refreshes without having to
> restart coordinator.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
