You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@hbase.apache.org by "stack (JIRA)" <ji...@apache.org> on 2018/10/17 18:48:00 UTC

[jira] [Resolved] (HBASE-21324) [balancer] Gives up too soon when lots of regions

     [ https://issues.apache.org/jira/browse/HBASE-21324?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

stack resolved HBASE-21324.
---------------------------
    Resolution: Not A Problem

Resolving as not a problem.  The balancer needs to be configured appropriately to run against a big cluster. Upping timeouts and max steps as well as downing per region steps got the balancer to make progress. That this config should be dynamic not requiring a restart and that the logging on balancer gives little indication that reconfig is needed are other issues.

> [balancer] Gives up too soon when lots of regions
> -------------------------------------------------
>
>                 Key: HBASE-21324
>                 URL: https://issues.apache.org/jira/browse/HBASE-21324
>             Project: HBase
>          Issue Type: Bug
>          Components: Balancer
>            Reporter: stack
>            Assignee: stack
>            Priority: Major
>
> When there a few regions, the balancer seems to do a poorish job. My cluster is all out of whack and while the balancer runs, it doesn't do anything. So, I need to configure the balancer to run against large number of regions. There is no doc in reguide to help and there is no logging coming out of the balancer to help either. Let me fix.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)