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

Posted to issues@hbase.apache.org by "Peter Somogyi (Jira)" <ji...@apache.org> on 2023/06/27 14:27:00 UTC

[jira] [Reopened] (HBASE-27892) Report memstore on-heap and off-heap size as jmx metrics

     [ https://issues.apache.org/jira/browse/HBASE-27892?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Peter Somogyi reopened HBASE-27892:
-----------------------------------

This commit causes compilation failure on branch-2.4. Reopening for the revert.

> Report memstore on-heap and off-heap size as jmx metrics
> --------------------------------------------------------
>
>                 Key: HBASE-27892
>                 URL: https://issues.apache.org/jira/browse/HBASE-27892
>             Project: HBase
>          Issue Type: Improvement
>          Components: metrics
>            Reporter: Bryan Beaudreault
>            Assignee: Jing Yu
>            Priority: Major
>             Fix For: 2.6.0, 2.4.18, 2.5.6, 3.0.0-beta-1
>
>
> Currently we only report "memStoreSize" metric in sub=RegionServer bean. I've noticed a big discrepancy between this metric and the RS UI's "Memstore On-Heap Size". It seems like "memStoreSize" is the overall data size, while the on-heap size is coming from our heap estimation which includes POJO heap overhead, etc.
> I have a regionserver with only 750mb of "memStoreSize", but the on-heap size is over 1gb.  This is non-trivial for estimating overall heap size necessary for a regionserver. Since we have the data, let's report it.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)