hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HDFS-1112) Edit log buffer should not grow unboundedly
Date Tue, 18 May 2010 07:01:49 GMT

    [ https://issues.apache.org/jira/browse/HDFS-1112?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12868551#action_12868551
] 

Todd Lipcon commented on HDFS-1112:
-----------------------------------

Some comments:
- timeToSync() method name implies that it's a time based condition, perhaps better to rename
to something like isBufferFull() since that's the policy currently implemented. Or more generically
shouldForceSync()
- why the notifyAll calls inside logSync, since that's done already inside doneWithAutoSyncScheduling?
- the changes to add IOException thrown seems messy - isn't our contract always that failing
to append to the edit log is a fatal error for the NN and thus a checked exception won't be
thrown?
- the patch doesn't apply to trunk anymore

Aside from that, think this looks good. I edited TestFSEditLogRace to run for a couple minutes
with 16 threads making edits simultaneously, and no deadlocks or anything.

> Edit log buffer should not grow unboundedly
> -------------------------------------------
>
>                 Key: HDFS-1112
>                 URL: https://issues.apache.org/jira/browse/HDFS-1112
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: name-node
>    Affects Versions: 0.22.0
>            Reporter: Hairong Kuang
>            Assignee: Hairong Kuang
>             Fix For: 0.22.0
>
>         Attachments: editLogBuf.patch
>
>
> Currently HDFS does not impose an upper limit on the edit log buffer. In case there are
a large number of open operations coming in with access time update on, since open does not
call sync automatically, there is a possibility that the buffer grow to a large size, therefore
causes memory leak and full GC in extreme cases as described in HDFS-1104. 
> The edit log buffer should be automatically flushed when the buffer becomes full.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message