2

OS : Windows

App Server: Wildfly 10

Java Version: 8

We have our application running in a remote environment isolated from the internet connection (it completely offline). It will run continuously until any new patch delivered to that server. In that case, we often facing the problem with ActiveMQ Artemis server (with in Wildfly 10) which is getting unreachable from the application code seems like it's getting shutdown. But note that Undertow server is fine able to access application features.

Observed that if the server is continuously running for more than 30 days then the problem is occurring. At present to resolve the issue we have been instructed to restart the apps in a frequency. But looking for the real root cause and how to resolve it.

Exception:

javax.jms.JMSException: Failed to create session factory
at org.apache.activemq.artemis.jms.client.ActiveMQConnectionFactory.createConnectionInternal(ActiveMQConnectionFactory.java:727)
at org.apache.activemq.artemis.jms.client.ActiveMQConnectionFactory.createQueueConnection(ActiveMQConnectionFactory.java:284)
at 
Caused by: ActiveMQNotConnectedException[errorType=NOT_CONNECTED message=AMQ119007: Cannot connect to server(s). Tried with all available servers.]
at org.apache.activemq.artemis.core.client.impl.ServerLocatorImpl.createSessionFactory(ServerLocatorImpl.java:778)
at org.apache.activemq.artemis.jms.client.ActiveMQConnectionFactory.createConnectionInternal(ActiveMQConnectionFactory.java:724)
... 71 more
Justin Bertram
  • 29,372
  • 4
  • 21
  • 43
Jeevanantham
  • 984
  • 3
  • 19
  • 48
  • 1
    Is there anything in the Wildfly log that would indicate trouble with the Artemis instance? Can you verify network connectivity isn't a problem? – Justin Bertram Jul 05 '19 at 16:48
  • 1
    Verified log, Issue is not related to network connectivity. Application is using the wildfly inbuild ActiveMQ service so hope the network is not the issue – Jeevanantham Jul 09 '19 at 09:14
  • This seems like an environmental issue to me since everything works fine for 30 days and then just stops working and you've provided no evidence that the Artemis instance has actually stopped. – Justin Bertram Aug 15 '19 at 13:05
  • Are u sure artemis broker already starting up? check wildfly startup log to verify artemis broker already starting. – Awan Biru Aug 22 '19 at 04:21
  • @AwanBiru it's starting up and process messages and suddendly some interval of time issues pops up – Jeevanantham Sep 26 '19 at 13:55

2 Answers2

0

If possible, try with a separate activeMQ server.

It may be some memory leak issues as wildfly built-in may have some heap not reclaimed.

The only way to make sure is to check the memory usage logs and/or start wildfly with more/less memory to see if your 30 days is increased or reduced with memory.

Edward Aung
  • 3,014
  • 1
  • 12
  • 15
0

Not enough information, perhaps DEBUG level log would reveal more info. This error was thrown by Artemis JMS client implementation(createConnectionInternal). Calls to AMQ broker failed perhaps because of:-

  1. Artemis broker does not starting up. Verify server log to ensure broker starting up properly

  2. The messaging-subsystem and connector configuration problem. Verify Artemis subsystem configuration at standalone.xml or domain.xml. Best work with jboss admin to ensure all messaging subsystem is correct and working.

Awan Biru
  • 373
  • 2
  • 10