[
https://issues.apache.org/jira/browse/HBASE-13496?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-13496:
--------------------------
Attachment: onheapoffheapcompare.tgz
My little JMH project after following these instructions:
http://openjdk.java.net/projects/code-tools/jmh/ I still need to learn it but
seems very nice for this kinda thing. +1 on using this going forward (lets
think about how we'd add these to our codebase; its a little awkward in that
jmh wants to own the parent pom and running the test is done against a built
jar).
I made two 'benchmarks' out of [~anoop.hbase]'s class unrolling the loop (jmh
warns about how loops can color measurement). One using onheap compare and
another offheap with unsafe compares. I then ran one test and then another
using flags to specify forks and iterations. I ran them independently (see
above for how I did it).
> Make Bytes$LexicographicalComparerHolder$UnsafeComparer::compareTo inlineable
> -----------------------------------------------------------------------------
>
> Key: HBASE-13496
> URL: https://issues.apache.org/jira/browse/HBASE-13496
> Project: HBase
> Issue Type: Sub-task
> Components: Scanners
> Reporter: Anoop Sam John
> Assignee: Anoop Sam John
> Fix For: 2.0.0, 1.2.0
>
> Attachments: ByteBufferUtils.java, HBASE-13496.patch,
> OffheapVsOnHeapCompareTest.java, onheapoffheapcompare.tgz
>
>
> While testing with some other perf comparisons I have noticed that the above
> method (which is very hot in read path) is not getting inline
> bq.@ 16
> org.apache.hadoop.hbase.util.Bytes$LexicographicalComparerHolder$UnsafeComparer::compareTo
> (364 bytes) hot method too big
> We can do minor refactoring to make it inlineable.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
