[
https://issues.apache.org/jira/browse/HBASE-5898?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Elliott Clark updated HBASE-5898:
---------------------------------
Attachment: HBASE-5898-0.patch
We saw this on a production cluster.
A single data block was being requested by all ipc handlers. This caused a
great deal of lock contention. Todd's patch fixed the issue and caused reads
to go back to a normal duration.
This patch is almost identical to Todd's patch. It makes the second try to get
a block from the block cache not count as a miss.
> Consider double-checked locking for block cache lock
> ----------------------------------------------------
>
> Key: HBASE-5898
> URL: https://issues.apache.org/jira/browse/HBASE-5898
> Project: HBase
> Issue Type: Improvement
> Components: Performance
> Affects Versions: 0.94.1
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Priority: Critical
> Attachments: 5898-TestBlocksRead.txt, HBASE-5898-0.patch,
> hbase-5898.txt
>
>
> Running a workload with a high query rate against a dataset that fits in
> cache, I saw a lot of CPU being used in IdLock.getLockEntry, being called by
> HFileReaderV2.readBlock. Even though it was all cache hits, it was wasting a
> lot of CPU doing lock management here. I wrote a quick patch to switch to a
> double-checked locking and it improved throughput substantially for this
> workload.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
