You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ozone.apache.org by "Siddhant Sangwan (Jira)" <ji...@apache.org> on 2021/10/19 09:31:00 UTC
[jira] [Commented] (HDDS-5760) ContainerBalancer incorrectly exits
an iteration early without checking move results.
[ https://issues.apache.org/jira/browse/HDDS-5760?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17430436#comment-17430436 ]
Siddhant Sangwan commented on HDDS-5760:
----------------------------------------
HDDS-5671 fixes this by refactoring the end of iteration checks to a separate method, ContainerBalancer#checkIterationMoveResults.
> ContainerBalancer incorrectly exits an iteration early without checking move results.
> -------------------------------------------------------------------------------------
>
> Key: HDDS-5760
> URL: https://issues.apache.org/jira/browse/HDDS-5760
> Project: Apache Ozone
> Issue Type: Sub-task
> Components: SCM
> Reporter: Siddhant Sangwan
> Assignee: Siddhant Sangwan
> Priority: Major
>
> In ContainerBalancer#doIteration:
> {code}
> IterationResult result = checkConditionsForBalancing();
> if (result != null) {
> LOG.info("Exiting current iteration: {}", result);
> return result;
> }
> {code}
> Leads to the iteration ending early without checking move results and updating metrics. This bug can be reproduced by setting the {{maxSizeToMovePerIteration}} to a small value such as 5GB that allows one container move before ending the iteration.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org