activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Keith Oppenheim (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (AMQ-5133) Selector Worker thread consumes significant CPU when broker is idle
Date Thu, 03 Apr 2014 18:29:17 GMT

     [ https://issues.apache.org/jira/browse/AMQ-5133?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Keith Oppenheim updated AMQ-5133:
---------------------------------

    Description: The Selector Worker thread is consuming 25% of my host CPU even though there
are no active producers or consumers and there are no messages in the message store.  Unfortunately,
I don't know of way to reproduce this issue.  I was able to identify the offending thread
by first using {{ps -eLF}} to find the ActiveMQ thread with a relatively large and continually
increasing amount of CPU time.  I then correlated the thread ID (LWP  column  in ps output)
to the   (was: The Selector Worker thread is consuming 25% of my host CPU even though there
are no active producers or consumers and there are no messages in the message store.  Unfortunately,
I don't know of way to reproduce this issue.  I was able to identify the offending thread
by first using {{ps -eLF}} to find the ActiveMQ thread with a relatively large and continually
increasing amount of CPU time.  I then correlated the thread ID (LWP  )

> Selector Worker thread consumes significant CPU when broker is idle
> -------------------------------------------------------------------
>
>                 Key: AMQ-5133
>                 URL: https://issues.apache.org/jira/browse/AMQ-5133
>             Project: ActiveMQ
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: 5.9.0
>         Environment: OS:CentOS release 6.5
> Java: 1.6.0_17, Java HotSpot 64-Bit Server VM
> Host: VMWare Xeon(R) CPU E5-2660 0; 4 core @2.2GHz
>            Reporter: Keith Oppenheim
>
> The Selector Worker thread is consuming 25% of my host CPU even though there are no active
producers or consumers and there are no messages in the message store.  Unfortunately, I don't
know of way to reproduce this issue.  I was able to identify the offending thread by first
using {{ps -eLF}} to find the ActiveMQ thread with a relatively large and continually increasing
amount of CPU time.  I then correlated the thread ID (LWP  column  in ps output) to the 



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message