[
https://issues.apache.org/jira/browse/HADOOP-1272?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
dhruba borthakur updated HADOOP-1272:
-------------------------------------
Attachment: innerclasses4.patch
Hi Konstantin, thanks for the feedback. I agree that moving the SafeModeInfo
out of the FSNamesystem file makes the code layout has become a little more
involved that earlier. But the assumed simplicity that was earlier present was
because of global variables (global with respect to FSNamesystem). I think this
patch actually ensures that methods in SafeModeInfo cannot access private
members of FSNamesystem directly. This should improve code maintainability and
cleanliness.
The fact the SafeModeInfo cannot access private members of FSNamesystem means
that a log message that was earlier present in one place is now duplicated at
two or three places. I have removed the assert because SafeModeInfo cannot be
accessing private variables in FSNamesystem.
Thanks for catching the problems with javadoc. I have fixed them in this
version of this patch.
I would like to make this "patch available" because that might make other
people review this patch.
> Extract InnerClasses from FSNamesystem into separate classes
> ------------------------------------------------------------
>
> Key: HADOOP-1272
> URL: https://issues.apache.org/jira/browse/HADOOP-1272
> Project: Hadoop
> Issue Type: Bug
> Components: dfs
> Reporter: dhruba borthakur
> Assigned To: dhruba borthakur
> Attachments: innerclasses4.patch
>
>
> This will make the code cleaner. Also, it leads itself to a cleaner and
> easily understandable finer-grain locking model.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
