You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@activemq.apache.org by "Marcus Malcom (JIRA)" <ji...@apache.org> on 2009/07/28 21:29:34 UTC
[jira] Created: (AMQ-2333) Active MQ performance issues when there
are more than a 1000 queue'd up messages
Active MQ performance issues when there are more than a 1000 queue'd up messages
--------------------------------------------------------------------------------
Key: AMQ-2333
URL: https://issues.apache.org/activemq/browse/AMQ-2333
Project: ActiveMQ
Issue Type: Bug
Components: Broker, JMS client
Affects Versions: 5.2.0
Reporter: Marcus Malcom
Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
Our system pretty much has all the defaults.
Note: this was not a problem before upgrading to 5.2.0
Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (AMQ-2333) Active MQ performance issues when
there are more than 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=53047#action_53047 ]
Marcus Malcom commented on AMQ-2333:
------------------------------------
OK, I'm going to try the systemUsage first. Here my current config:
{panel}
<!-- The maximum about of space the broker will use before slowing down producers -->
<systemUsage>
<systemUsage>
<memoryUsage>
<memoryUsage limit="20 mb"/>
</memoryUsage>
<storeUsage>
<storeUsage limit="1 gb" name="foo"/>
</storeUsage>
<tempUsage>
<tempUsage limit="100 mb"/>
</tempUsage>
</systemUsage>
</systemUsage>
{panel}
Switching it to:
{panel}
<memoryUsage limit="50 mb"/>
<tempUsage limit="500 mb"/>
{panel}
Hopefully that will fix the problem - will let you know shortly.
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
> Priority: Critical
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (AMQ-2333) Active MQ performance issues when
there are more than a 1000 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=53024#action_53024 ]
Marcus Malcom commented on AMQ-2333:
------------------------------------
This just happened w/ only 142 messages in a queue
> Active MQ performance issues when there are more than a 1000 queue'd up messages
> --------------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (AMQ-2333) Active MQ performance issues when
there are more than 100 queue'd up messages
Posted by "Gary Tully (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=53031#action_53031 ]
Gary Tully commented on AMQ-2333:
---------------------------------
investigate turning off producer flow control and increasing the memory available to the broker through the systemUsage config.
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
> Priority: Critical
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (AMQ-2333) Active MQ performance issues when there
are more than a 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Marcus Malcom updated AMQ-2333:
-------------------------------
Summary: Active MQ performance issues when there are more than a 100 queue'd up messages (was: Active MQ performance issues when there are more than a 1000 queue'd up messages)
> Active MQ performance issues when there are more than a 100 queue'd up messages
> -------------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (AMQ-2333) Active MQ performance issues when there
are more than 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Marcus Malcom updated AMQ-2333:
-------------------------------
Priority: Critical (was: Major)
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
> Priority: Critical
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (AMQ-2333) Active MQ performance
issues when there are more than 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=53047#action_53047 ]
Marcus Malcom edited comment on AMQ-2333 at 7/30/09 8:23 AM:
-------------------------------------------------------------
OK, I'm going to try the systemUsage first. Here's my current config:
{panel}
<!-- The maximum about of space the broker will use before slowing down producers -->
<systemUsage>
<systemUsage>
<memoryUsage>
<memoryUsage limit="20 mb"/>
</memoryUsage>
<storeUsage>
<storeUsage limit="1 gb" name="foo"/>
</storeUsage>
<tempUsage>
<tempUsage limit="100 mb"/>
</tempUsage>
</systemUsage>
</systemUsage>
{panel}
Switching it to:
{panel}
<memoryUsage limit="50 mb"/>
<tempUsage limit="500 mb"/>
{panel}
Hopefully that will fix the problem - will let you know shortly.
was (Author: malcomm@ochin.org):
OK, I'm going to try the systemUsage first. Here my current config:
{panel}
<!-- The maximum about of space the broker will use before slowing down producers -->
<systemUsage>
<systemUsage>
<memoryUsage>
<memoryUsage limit="20 mb"/>
</memoryUsage>
<storeUsage>
<storeUsage limit="1 gb" name="foo"/>
</storeUsage>
<tempUsage>
<tempUsage limit="100 mb"/>
</tempUsage>
</systemUsage>
</systemUsage>
{panel}
Switching it to:
{panel}
<memoryUsage limit="50 mb"/>
<tempUsage limit="500 mb"/>
{panel}
Hopefully that will fix the problem - will let you know shortly.
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
> Priority: Critical
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (AMQ-2333) Active MQ performance issues when there
are more than 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Marcus Malcom updated AMQ-2333:
-------------------------------
Summary: Active MQ performance issues when there are more than 100 queue'd up messages (was: Active MQ performance issues when there are more than a 100 queue'd up messages)
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (AMQ-2333) Active MQ performance issues when there
are more than 100 queue'd up messages
Posted by "Marcus Malcom (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/activemq/browse/AMQ-2333?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Marcus Malcom resolved AMQ-2333.
--------------------------------
Resolution: Fixed
The configuration changes that I made fixed our problem. Thank you for the help!
> Active MQ performance issues when there are more than 100 queue'd up messages
> -----------------------------------------------------------------------------
>
> Key: AMQ-2333
> URL: https://issues.apache.org/activemq/browse/AMQ-2333
> Project: ActiveMQ
> Issue Type: Bug
> Components: Broker, JMS client
> Affects Versions: 5.2.0
> Reporter: Marcus Malcom
> Priority: Critical
>
> Over the past couple of days some of our queues get rather full because of downstream problems. The messages start numbering in the 1000's. When that happens ActiveMQ slows way down. I believe is slows down because we are trying to produce a message to the overloaded queue and it's taking a long time (minutes instead of seconds). Once the overloaded queue is emptied the problems go away.
> Our system pretty much has all the defaults.
> Note: this was not a problem before upgrading to 5.2.0
> Any ideas on what should be done?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.