[
https://issues.apache.org/jira/browse/HBASE-17017?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15655663#comment-15655663
]
Mikhail Antonov commented on HBASE-17017:
-----------------------------------------
[~enis] the only thing i had for 1.3RC was HBASE-16852 that I discovered to be
flaky, and as this is compaction related I'm looking at it now, since we've had
enough regressions in compactions lately; But since the win as demonstrated is
sizable, and we can remove metrics in 1.3.0 (and if we want to and if faster
impl is available we can bring some back in 1.3.x), but we won't be able to
remove metrics in 1.3.1, I'd be +1 for that to go in 1.3.0.
One thing - your benchmark is from master, right? Implementation of histograms
on branch-1 is more expensive I believe (no LongAdder), so I'd be really
curious if you could repeat your benchmark on branch-1.3?
> Remove the current per-region latency histogram metrics
> -------------------------------------------------------
>
> Key: HBASE-17017
> URL: https://issues.apache.org/jira/browse/HBASE-17017
> Project: HBase
> Issue Type: Sub-task
> Components: metrics
> Affects Versions: 2.0.0, 1.4.0
> Reporter: Duo Zhang
> Assignee: Enis Soztutar
> Fix For: 2.0.0, 1.4.0
>
> Attachments: Screen Shot 2016-11-04 at 3.00.21 PM.png, Screen Shot
> 2016-11-04 at 3.38.42 PM.png, hbase-17017_v1.patch, hbase-17017_v2.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
