- How can I configure the broker's memory limits
- How do the memory limits affect Producer Flow Control
- We have a number of routes that depend on ActiveMQ endpoints. We are finding that when our queues get to around 1,000 messages and we don't have a consumer that we end up with a large number of inflight exchanges and everything stops. As soon as we purge the queue everything recovers. Reading about ActiveMQ and looking at our logs we found what we thought was the reason Producer Flow Control. We have now modified the default activemq.xml configuration in the /etc directory of fuse to
<destinationPolicy> <policyMap> <policyEntries> <policyEntry topic=">" producerFlowControl="false"> </policyEntry> <policyEntry queue=">" producerFlowControl="false"> </policyEntry> </policyEntries> </policyMap> </destinationPolicy>
In the belief that it would stop this from occurring but it doesn't.
- Are there any known issues with delayed cleanup / memory leaks around using the Camel JMS component and ActiveMQ with the failover transport?We appear to have a situation where ActiveMQBytesMessage instances hang around in the client heap much longer than expected.
It appears some messages hang for 6-12 hours after the JMS producer has completed
It does not seem to happen with all messages
The messages seem to eventually get cleaned up
The messages vary in size from ~10KB to ~60MB, so sometimes we see OutOfMemoryErrors.
- continual climb in A-MQ Memory Percent Used. How to diagnose it?
- Fuse ESB Enterprise
- Fuse MQ Enterprise
- Fuse Message Broker
- Red Hat JBoss A-MQ
- Red Hat JBoss Fuse
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.