AMQ 7: Critical analyser cannot shut down a broker in out-of-memory situations
Issue
The AMQ 7 'critical analyser' is designed to shut down the broker in certain error conditions, in the expectation that this will reduce the likelihood of data corruption. So you might expect a broker shutdown after a log message like this:
2023-06-07 12:51:02,114 ERROR [org.apache.activemq.artemis.core.server] AMQ224079: The process for the virtual machine will be killed, as component QueueImpl[name=foo, postOffice=PostOfficeImpl [server=ActiveMQServerImpl::name=bar], temp=false]@410a2f78 is not responsive
However, in low-memory situations the broker might not shut down. This is problematic in when some infrastructure is standing by to start a new broker instance when the previous one fails. Most likely, the fact that the broker is still running will prevent a new instance starting.
Environment
- Red Hat AMQ
- 7.x, all releases
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.