[
https://issues.apache.org/jira/browse/HBASE-7180?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13528677#comment-13528677
]
Lars Hofhansl commented on HBASE-7180:
--------------------------------------
Yeah... That's better name.
Sorry, should have mentioned the change to synchronized on some of the public
next(...) methods.
It was intentional... These all end up calling the same next method (which is
synchronized), so they did not need to be synchronized and we avoid double
synchronization (the performance gain from this will be minimal, but it is
better form - I think)
(I think at the RegionScanner level we could eventually remove all
synchronization - it does not make sense for client to interleave calls to
next() on the *same* scanner between multiple threads. But that's for another
issue.)
> RegionScannerImpl.next() is inefficient.
> ----------------------------------------
>
> Key: HBASE-7180
> URL: https://issues.apache.org/jira/browse/HBASE-7180
> Project: HBase
> Issue Type: Bug
> Reporter: Lars Hofhansl
> Assignee: Lars Hofhansl
> Fix For: 0.96.0, 0.94.4
>
> Attachments: 7180-0.94-SKETCH.txt, 7180-0.94-v1.txt,
> 7180-0.94-v2.txt, 7180-0.94-v3.txt, 7180-0.94-v4.txt, 7180-0.96-v1.txt,
> 7180-0.96-v2.txt, 7180-0.96-v3.txt
>
>
> We just came across a special scenario.
> For our Phoenix project (SQL runtime for HBase), we push a lot of work into
> HBase via coprocessors. One method is to wrap RegionScanner in coprocessor
> hooks and then do processing in the hook to avoid returning a lot of data to
> the client unnecessarily.
> In this specific case this is pretty bad. Since the wrapped RegionScanner's
> next() does not "know" that it is called this way is still does all of this
> on each invocation:
> # Starts a RegionOperation
> # Increments the request count
> # set the current read point on a thread local (because generally each call
> could come from a different thread)
> # Finally does the next on its StoreScanner(s)
> # Ends the RegionOperation
> When this is done in a tight loop millions of times (as is the case for us)
> it starts to become significant.
> Not sure what to do about this, really. Opening this issue for discussion.
> One way is to extend the RegionScanner with an "internal" next() method of
> sorts, so that all this overhead can be avoided. The coprocessor could call
> the regular next() methods once and then just call the cheaper internal
> version.
> Are there better/cleaner ways?
--
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
