[
https://issues.apache.org/jira/browse/MAPREDUCE-7332?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17308876#comment-17308876
]
Ahmed Hussein commented on MAPREDUCE-7332:
------------------------------------------
To be on the safe side, I preferred to use the guava library which should be
compatible for both JDK7 and JDK8.
To achieve the same level of concurrency, I used {{LoadingCache}} which
guarantees atomicity of the operations.
I experimented {{guava.MapMaker().makeMap()}} but the unit test
{{TestMRIntermediateDataEncryption}} --- with multiple reducers -- failed.
The failure was due to the fact that the {{ConcurrentMap}} returned from
{{guava.MapMaker().makeMap()}} did not guarantee atomicity.
I tested the patch locally and reviewed the checkstyle.
{code:bash}
dev-support/bin/test-patch \
--plugins="maven,checkstyle" \
--test-parallel=true \
--branch=branch-2.10 \
patches/MAPREDUCE-7332.branch-2.10.001.patch
{code}
> Fix SpillCallBackPathsFinder to use JDK7 on branch-2.10
> -------------------------------------------------------
>
> Key: MAPREDUCE-7332
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-7332
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: job submission, security
> Reporter: Ahmed Hussein
> Assignee: Ahmed Hussein
> Priority: Minor
> Attachments: MAPREDUCE-7332.branch-2.10.001.patch
>
>
> I mistakenly uploaded a patch for branch-2.10 that uses JDK8.
> Yetus did not fail though. It should be investigated why it was not failing
> if JDK8+ is used in the code.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
