[
https://issues.apache.org/jira/browse/HBASE-14268?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14730145#comment-14730145
]
Hiroshi Ikeda commented on HBASE-14268:
---------------------------------------
I measured it just now in my environment, which has 8 cpu core, and the result
is as followed. For each KeyLocker I run the performance application 5 times
and I collected the output number which represents the consumed time
(nanoseconds).
old KeyLocker
(1) 4168557944
(2) 4173845279
(3) 4276035366
(4) 4344219315
(5) 4393414763
average
= 4271214533.4 (nanos)
~ 4.27 (sec)
new KeyLocker
(1) 270832002
(2) 318058811
(3) 278171946
(4) 265603446
(5) 279867215
average
= 282506684 (nanos)
~ 0.28 (sec)
The difference just comes from overhead of context switches when
getting/returning locks from/to the pool, and this may be not so meaningful for
heavy tasks which cause context switches.
> Improve KeyLocker
> -----------------
>
> Key: HBASE-14268
> URL: https://issues.apache.org/jira/browse/HBASE-14268
> Project: HBase
> Issue Type: Improvement
> Components: util
> Reporter: Hiroshi Ikeda
> Assignee: Hiroshi Ikeda
> Priority: Minor
> Attachments: 14268-V5.patch, HBASE-14268-V2.patch,
> HBASE-14268-V3.patch, HBASE-14268-V4.patch, HBASE-14268-V5.patch,
> HBASE-14268-V5.patch, HBASE-14268-V6.patch, HBASE-14268-V7.patch,
> HBASE-14268-V7.patch, HBASE-14268.patch, KeyLockerPerformance.java
>
>
> 1. In the implementation of {{KeyLocker}} it uses atomic variables inside a
> synchronized block, which doesn't make sense. Moreover, logic inside the
> synchronized block is not trivial so that it makes less performance in heavy
> multi-threaded environment.
> 2. {{KeyLocker}} gives an instance of {{RentrantLock}} which is already
> locked, but it doesn't follow the contract of {{ReentrantLock}} because you
> are not allowed to freely invoke lock/unlock methods under that contract.
> That introduces a potential risk; Whenever you see a variable of the type
> {{RentrantLock}}, you should pay attention to what the included instance is
> coming from.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
