[
https://issues.apache.org/jira/browse/TINKERPOP-1863?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16316666#comment-16316666
]
ASF GitHub Bot commented on TINKERPOP-1863:
-------------------------------------------
Github user spmallette commented on the issue:
https://github.com/apache/tinkerpop/pull/771
Thanks @EugeneChung - as this was minor change that was discussed on the
dev mailing list already I decided to just CTR and merge it without further
votes. I updated the CHANGELOG and made a minor tweak to the equality check for
null:
https://github.com/apache/tinkerpop/commit/c4b0f9de04fda08db5b6b03a91ba58f3def0217a
> Delaying the setting of requestId till the RequestMessage instantiation time
> ----------------------------------------------------------------------------
>
> Key: TINKERPOP-1863
> URL: https://issues.apache.org/jira/browse/TINKERPOP-1863
> Project: TinkerPop
> Issue Type: Improvement
> Components: driver
> Reporter: Eugene Chung
> Priority: Minor
>
> The Builder class of
> org.apache.tinkerpop.gremlin.driver.message.RequestMessage class sets its
> requestId field as UUID.randomUUID() by default.
> But I think it should be fixed not to be set by default. The reasons are
> below;
> - UUID.randomUUID() uses SecureRandom which grabs the lock at JVM level,
> which means whole threads calling this API compete against each other.
> - Getting random value from SecureRandom is somewhat CPU-intensive job.
> - If a gremlin client sends requestId by itself, the costs above are useless.
> https://lists.apache.org/thread.html/a7e6cfeadad10852a2deed8616e47df6738c13c47119c12f98dcd3e1@%3Cdev.tinkerpop.apache.org%3E
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
