Jerico Sandhorn
Jerico Sandhorn

Reputation: 2020

Why does ActiveMQ restart automatically and how do i prevent it?

We've been using AMQ 5.5.1 in production for several months. Occasionally, we observe that the broker decides to refresh itself with no outside trigger. When this happens, our queue senders fail until the broker is back online some 10 minutes later. I cannot find any information or settings that would cause this behavior .. and let me control it.

Is this normal for the broker to recycle on its own like this? If so, what things would cause it?

2012-12-11 11:02:11,603 | INFO  | Refreshing   org.apache.activemq.xbean.XBeanBrokerFactory$1@f6ac0b: startup date [Tue Dec 11 11:02:11 EST 2012]; root of context hierarchy | org.apache.activemq.xbean.XBeanBrokerFactory$1 | WrapperSimpleAppMain
 2012-12-11 11:02:13,806 | WARN  | destroyApplicationContextOnStop parameter is deprecated, please use shutdown hooks instead | org.apache.activemq.xbean.XBeanBrokerService | WrapperSimpleAppMain
 2012-12-11 11:02:13,821 | INFO  | PListStore:D:\Tools\ActiveMQ\apache-activemq-5.5.1\bin\win32\..\..\data\localhost\tmp_storage started | org.apache.activemq.store.kahadb.plist.PListStore | WrapperSimpleAppMain
 2012-12-11 11:02:13,868 | INFO  | Using Persistence Adapter: KahaDBPersistenceAdapter[D:\Tools\ActiveMQ\apache-activemq-5.5.1\bin\win32\..\..\data\kahadb] | org.apache.activemq.broker.BrokerService | WrapperSimpleAppMain 
 2012-12-11 11:02:16,618 | INFO  | KahaDB is version 3 | org.apache.activemq.store.kahadb.MessageDatabase | WrapperSimpleAppMain
 2012-12-11 11:02:16,697 | INFO  | Recovering from the journal ... | org.apache.activemq.store.kahadb.MessageDatabase | WrapperSimpleAppMain

Upvotes: 3

Views: 3205

Answers (1)

Jerico Sandhorn
Jerico Sandhorn

Reputation: 2020

I found that the wrapper exe process was forcing the restart.

I was able to see in the wrapper.log (windows service) that the process was being restarted because the JVM was not responding. So this is not an issue with the broker auto-restarting per se.. it was an issue with the broker JVM somehow becoming hung (separate problem).

Here are the wrapper log entries for those interested:

ERROR  | wrapper  | 2012/12/11 11:01:58 | JVM appears hung: Timed out waiting for signal from JVM.
ERROR  | wrapper  | 2012/12/11 11:01:58 | JVM did not exit on request, terminated
STATUS | wrapper  | 2012/12/11 11:02:04 | Launching a JVM...

Upvotes: 1

Related Questions