[
https://issues.apache.org/jira/browse/ARTEMIS-3619?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17463063#comment-17463063
]
Dave Myron commented on ARTEMIS-3619:
-------------------------------------
It eventually stopped creating new .cmp files and renamed them all to .amq
files. Now it has almost 400GB of journal files for 214k scheduled messages.
Something is very wrong…
> Compacting generates .cmp files until disk is full when scheduled messages
> are in journal files
> -----------------------------------------------------------------------------------------------
>
> Key: ARTEMIS-3619
> URL: https://issues.apache.org/jira/browse/ARTEMIS-3619
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
> Affects Versions: 2.18.0
> Reporter: Dave Myron
> Priority: Critical
>
> We use the default of 10MB for journal files. When we schedule a couple
> hundred messages / jobs for a later time, Artemis begins creating 10MB .cmp
> files (like activemq-data-455341.amq.cmp) at a rate of about one hundred per
> minute until the disk is full and the server shuts down.
> Upon restarting, often a couple hours of CPU time is required to get back to
> a working state.
> Nothing is in the logs.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
