[
https://issues.apache.org/jira/browse/HDFS-5561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13832343#comment-13832343
]
Haohui Mai commented on HDFS-5561:
----------------------------------
I'm okay with either way. The original patch, however, only calls
loggers.size() once instead of every iteration.
You might argue that a trace-based JIT can inline the call and optimize it
away, but this depends on which JRE / JDK you're using.
> FSNameSystem#getNameJournalStatus() in JMX should return plain text instead
> of HTML
> -----------------------------------------------------------------------------------
>
> Key: HDFS-5561
> URL: https://issues.apache.org/jira/browse/HDFS-5561
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Affects Versions: 3.0.0, 2.2.0
> Reporter: Fengdong Yu
> Assignee: Haohui Mai
> Priority: Minor
> Attachments: HDFS-5561.000.patch, NNUI.PNG
>
>
> Currently FSNameSystem#getNameJournalStatus() returns the status of the
> quorum stream, which is an HTML string. This should not happen since that
> getNameJournalStatus() is a JMX call. This will confuse the downstream
> clients (e.g., the web UI) and lead to incorrect result.
> This jira proposes to change the information to plain text.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
