[
https://issues.apache.org/jira/browse/AMQ-4119?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Timothy Bish resolved AMQ-4119.
-------------------------------
Resolution: Fixed
Fix Version/s: 5.8.0
Fixed on trunk.
> Add memory usage *count* to destination MBean for better detection of memory
> leaks
> ----------------------------------------------------------------------------------
>
> Key: AMQ-4119
> URL: https://issues.apache.org/jira/browse/AMQ-4119
> Project: ActiveMQ
> Issue Type: New Feature
> Components: JMX
> Affects Versions: 5.7.0
> Reporter: Stirling Chow
> Fix For: 5.8.0
>
> Attachments: AMQ4119.patch
>
>
> We've been doing a lot of tuning of memory usage limits and exploring the
> behaviour of producer flow control when these limits are reached. In so
> doing, we've wanted to validate that memory usage counts are not leaking
> (e.g., as a result of bugs like AMQ-4116).
> In particular, we want to verify that after a large flood of messages is
> processed and queues become empty, the memory usage on each queue resets to 0.
> The existing MBean for queues/topics shows "MemoryPercentUsage" but not the
> actual bytes that the queue thinks it is using. For large limits, the
> MemoryPercentUsage can report 0% while there is actually some memory usage
> remaining. In this case, it is difficult to detect leaks without running
> many test iterations until the MemoryPercentUsage increments.
> Knowing the actual byte count of memory usage is also a more direct way of
> determining appropriate limits, since these limits themselves are expressed
> in bytes.
> Attached is a simple patch that adds a "MemoryUsage" attribute to the
> destination MBean.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
