[
https://issues.apache.org/jira/browse/HBASE-14712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14979332#comment-14979332
]
Matteo Bertozzi commented on HBASE-14712:
-----------------------------------------
done a quick test on branch-1.2
start cluster, stop cluster, start cluster some admin, stop cluster, more
admin... but I see the files going away.
quick look at the code the only place where we accumulate logs is when the log
is corrupted... (there is a TODO there). but you should see a "got an exception
while reading the procedure WAL" do you see it?
the other case is if you are hammering the admin operations and your default
expire time for ack requests is 5min or 15min in case the client crashes... you
may accumulate files for 5/15min but then they should go away.
so.. I go back thinking that you probably have a corruption? can you check
[~eclark]
> MasterProcWALs never clean up
> -----------------------------
>
> Key: HBASE-14712
> URL: https://issues.apache.org/jira/browse/HBASE-14712
> Project: HBase
> Issue Type: Bug
> Reporter: Elliott Clark
> Assignee: Matteo Bertozzi
> Priority: Blocker
>
> MasterProcWALs directory grows pretty much un-bounded. Because of that when
> master failover happens the NN is flooded with connections and everything
> grinds to a halt.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
