[
https://issues.apache.org/jira/browse/HBASE-15160?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15240975#comment-15240975
]
Enis Soztutar commented on HBASE-15160:
---------------------------------------
Thanks [~carp84] for the update. I have a half-finished patch for the next
version of this. Mind if I finish that, and you can do a review?
bq. Benchmarking with the JMH case Enis wrote up we could see using
MutableHistogram is less expensive thus promising.
BTW, I could not update the parent issue. The JHM test using the mutable
histograms was done against the mutable histogram implementation from
HBase-1.2.0. This is Hadoop's slow histograms. HBASE-15222 makes a huge
difference in terms of the micro-benchmark numbers, so the comparison between
inline histogram update versus ArrayBlockingQueue is even higher.
> Put back HFile's HDFS op latency sampling code and add metrics for monitoring
> -----------------------------------------------------------------------------
>
> Key: HBASE-15160
> URL: https://issues.apache.org/jira/browse/HBASE-15160
> Project: HBase
> Issue Type: Sub-task
> Affects Versions: 2.0.0, 1.1.2
> Reporter: Yu Li
> Assignee: Yu Li
> Attachments: HBASE-15160.patch, HBASE-15160_v2.patch,
> HBASE-15160_v3.patch
>
>
> In HBASE-11586 all HDFS op latency sampling code, including fsReadLatency,
> fsPreadLatency and fsWriteLatency, have been removed. There was some
> discussion about putting them back in a new JIRA but never happened.
> According to our experience, these metrics are useful to judge whether issue
> lies on HDFS when slow request occurs, so we propose to put them back in this
> JIRA, and add the metrics for monitoring as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
