activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chip Stettler (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMQ-5485) kahadb size grows continually with no queue activity
Date Mon, 15 Dec 2014 16:16:13 GMT

    [ https://issues.apache.org/jira/browse/AMQ-5485?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14246788#comment-14246788
] 

Chip Stettler commented on AMQ-5485:
------------------------------------

I attached my activemq config xml.  

As I was sending this it hit me as to what maybe going on.  I dont know the internals of ActiveMQ
or KahaDB, but it appears that one of the two is writing something to the DB even when idle.
 If this statement is true then it also makes sense that the size will increase until a journal
file is  filled up.  I had mine set to 100kb orginally.  As a test I set the journal size
to 1KB.  Now is see the the initialize size was something like 54K.  It increased to 56K but
never went beyound that.  It grows and shrinks betweeen 54K and 56K.



> kahadb size grows continually with no queue activity
> ----------------------------------------------------
>
>                 Key: AMQ-5485
>                 URL: https://issues.apache.org/jira/browse/AMQ-5485
>             Project: ActiveMQ
>          Issue Type: Bug
>    Affects Versions: 5.9.0
>            Reporter: Chip Stettler
>         Attachments: upf-activemq-server-config.xml
>
>
> We are having issues where the storage max size is reached.  After watching the jconsole
the kahadb size is growing continually. I turned on some trace output and see that with each
Checkpoint the size increases.
> Is there a missing setting or a bug?
> 11 Dec 2014 13:31:00,339 DEBUG MessageDatabase: Checkpoint started.
> 11 Dec 2014 13:31:00,343 DEBUG MessageDatabase: Checkpoint done.
> 11 Dec 2014 13:31:05,365 DEBUG MessageDatabase: Checkpoint started.
> 11 Dec 2014 13:31:05,369 TRACE MessageDatabase: Last update: 1:10975, full gc candidates
set: [1]
> 11 Dec 2014 13:31:05,370 TRACE MessageDatabase: gc candidates after producerSequenceIdTrackerLocation:1,
[]
> 11 Dec 2014 13:31:05,370 TRACE MessageDatabase: gc candidates after ackMessageFileMapLocation:1,
[]
> 11 Dec 2014 13:31:05,370 TRACE MessageDatabase: gc candidates after tx range:[null, null],
[]
> 11 Dec 2014 13:31:05,370 TRACE MessageDatabase: gc candidates: []
> 11 Dec 2014 13:31:05,370 DEBUG MessageDatabase: Checkpoint done.
> 11 Dec 2014 13:31:10,390 DEBUG MessageDatabase: Checkpoint started.
> 11 Dec 2014 13:31:10,393 DEBUG MessageDatabase: Checkpoint done.
> 11 Dec 2014 13:31:15,412 DEBUG MessageDatabase: Checkpoint started.
> 11 Dec 2014 13:31:15,415 DEBUG MessageDatabase: Checkpoint done.
> 11 Dec 2014 13:31:20,425 DEBUG MessageDatabase: Checkpoint started.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message