[
https://issues.apache.org/jira/browse/HBASE-27617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17692990#comment-17692990
]
tianhang tang commented on HBASE-27617:
---------------------------------------
Hi [~bbeaudreault], thank you for sharing your thoughts!
{quote}
So once a table moves off a server, all of the table metrics should quickly
become 0.
{quote}
Correct. But another problem is that we can't judge whether the table is no
longer there, or the user's request is uneven. Of course we can easily draw
conclusions from other dimensions, but after all, I think it is not good enough
now.
{quote}
It seems like we could do the HashMap lookup once when region opens
{quote}
I like your idea! As I said before, I also think we should do some refactoring
of tableMetrics. And your idea should work.
But this patch itself isn't complicated and doesn't have any negative impact on
stability/performance/resource overhead. Considering the complexity of
refactoring, I think it is worth to merge at first. Then we can open another
issue for overall refactoring.
What do you think? Thank you.
> Make metrics from MetricsTableLatenciesImpl removable
> -----------------------------------------------------
>
> Key: HBASE-27617
> URL: https://issues.apache.org/jira/browse/HBASE-27617
> Project: HBase
> Issue Type: Bug
> Reporter: tianhang tang
> Assignee: tianhang tang
> Priority: Major
>
> Now the table latency metrics will last forever on a RS.
> Could reproduce this issue easily:
> 1. Create a test table and the region opened on RS_A;
> 2. Do a query;
> 3. Move the table to another group;
> 4. See JMX on RS_A, table latency metrics still exists and will not deleted.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
