[
https://issues.apache.org/jira/browse/HDFS-5276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13794299#comment-13794299
]
Colin Patrick McCabe commented on HDFS-5276:
--------------------------------------------
As Binglin pointed out, you cannot use volatile here because of the "lost
updates" problem. There is no way to atomically add one to a volatile
variable-- someone else may have updated it in between the time you read the
value, and the time you wrote back the value plus one.
I also suspect that just using volatile would still be slower than the
thread-local approach. The fundamental problem here is inter-CPU communication
is slow, and just substituting volatile for atomic doesn't solve that.
Thanks for the benchmarks, Binglin.
> FileSystem.Statistics got performance issue on multi-thread read/write.
> -----------------------------------------------------------------------
>
> Key: HDFS-5276
> URL: https://issues.apache.org/jira/browse/HDFS-5276
> Project: Hadoop HDFS
> Issue Type: Bug
> Affects Versions: 2.0.4-alpha
> Reporter: Chengxiang Li
> Assignee: Colin Patrick McCabe
> Attachments: DisableFSReadWriteBytesStat.patch, HDFS-5276.001.patch,
> HDFS-5276.002.patch, HDFSStatisticTest.java, hdfs-test.PNG, jstack-trace.PNG,
> TestFileSystemStatistics.java, ThreadLocalStat.patch
>
>
> FileSystem.Statistics is a singleton variable for each FS scheme, each
> read/write on HDFS would lead to a AutomicLong.getAndAdd(). AutomicLong does
> not perform well in multi-threads(let's say more than 30 threads). so it may
> cause serious performance issue. during our spark test profile, 32 threads
> read data from HDFS, about 70% cpu time is spent on
> FileSystem.Statistics.incrementBytesRead().
--
This message was sent by Atlassian JIRA
(v6.1#6144)
