[
https://issues.apache.org/jira/browse/HBASE-20942?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16595460#comment-16595460
]
Josh Elser commented on HBASE-20942:
------------------------------------
bq. I think the latter has worse outcomes from the former, would you not agree?
I don't think I am asking for too much here.
I honestly don't know what the right outcome is. I don't like the idea of
spinning out X Jira issues just to get a single change committed (the hoops to
get something committed everywhere is already pedantic). My plan of action will
be to just spin out clones for all branches requiring RM-response going
forward. I think this will avoid causing more heartburn for the situation you
described.
> Improve RpcServer TRACE logging
> -------------------------------
>
> Key: HBASE-20942
> URL: https://issues.apache.org/jira/browse/HBASE-20942
> Project: HBase
> Issue Type: Task
> Components: Operability
> Affects Versions: 2.1.0, 1.4.6
> Reporter: Esteban Gutierrez
> Assignee: Krish Dey
> Priority: Major
> Fix For: 3.0.0, 1.5.0, 1.3.3, 2.2.0, 2.0.2, 1.4.7
>
> Attachments: HBASE-20942.002.patch, HBASE-20942.003.patch,
> HBASE-20942.004.patch, HBASE-20942.005.patch
>
>
> Two things:
> * We truncate RpcServer output to 1000 characters for trace logging. Would
> be better if that value was configurable.
> * There is the chance for an ArrayIndexOutOfBounds when truncating the TRACE
> log message.
> Esteban mentioned this to me earlier, so I'm crediting him as the reporter.
> cc: [~elserj]
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
