Bamboo fails with error 'avax.jms.JMSException: Could not create Transport' after turning on SSO

[HN]

I notice the following errors crashing Bamboo after enabling SSO:

bamboo-0 bamboo 2018-04-18 00:45:24,368 ERROR [buildTailMessageListenerConnector-1] [BrokerService] Failed to start Apache ActiveMQ ([bamboo, null], java.lang.IllegalStateException: Shutdown in progress)
bamboo-0 bamboo 2018-04-18 00:45:24,404 ERROR [buildTailMessageListenerConnector-1] [FingerprintMatchingMessageListenerContainer] Could not refresh JMS Connection for destination 'queue://com.atlassian.bamboo.buildTailQueue' - retrying using FixedBackOff{interval=5000, currentAttempts=0, maxAttempts=unlimited}. Cause: Error while attempting to retrieve a connection from the pool; nested exception is javax.jms.JMSException: Could not create Transport. Reason: javax.jms.JMSException: Could not create Transport. Reason: java.lang.IllegalStateException: Shutdown in progress
bamboo-0 bamboo 2018-04-18 00:45:24,430 ERROR [ActiveMQ Connection Executor: vm://bamboo#0] [BrokerService] Failed to start Apache ActiveMQ ([bamboo, null], java.lang.IllegalStateException: Shutdown in progress)
bamboo-0 bamboo 2018-04-18 00:45:24,445 ERROR [ActiveMQ Connection Executor: vm://bamboo#0] [BambooServerSimpleMessageListenerContainer] Encountered non-recoverable JMSException
bamboo-0 bamboo 2018-04-18 00:45:24,461 ERROR [bambooAgentMessageListenerConnector-1] [BrokerService] Failed to start Apache ActiveMQ ([bamboo, null], java.lang.IllegalStateException: Shutdown in progress)
bamboo-0 bamboo 2018-04-18 00:45:24,475 ERROR [bambooAgentMessageListenerConnector-1] [FingerprintMatchingMessageListenerContainer] Could not refresh JMS Connection for destination 'queue://com.atlassian.bamboo.serverQueue' - retrying using FixedBackOff{interval=5000, currentAttempts=0, maxAttempts=unlimited}. Cause: Error while attempting to add new Connection to the pool; nested exception is javax.jms.JMSException: Could not create Transport. Reason: javax.jms.JMSException: Could not create Transport. Reason: java.lang.IllegalStateException: Shutdown in progress
bamboo-0 bamboo 2018-04-18 00:45:24,488 ERROR [bambooHeartBeatMessageListenerConnector-1] [BrokerService] Failed to start Apache ActiveMQ ([bamboo, null], java.lang.IllegalStateException: Shutdown in progress)
bamboo-0 bamboo 2018-04-18 00:45:24,506 ERROR [bambooHeartBeatMessageListenerConnector-1] [BambooDefaultMessageListenerContainer] Could not refresh JMS Connection for destination 'queue://com.atlassian.bamboo.heartbeatQueue' - retrying using FixedBackOff{interval=5000, currentAttempts=0, maxAttempts=unlimited}. Cause: Error while attempting to add new Connection to the pool; nested exception is javax.jms.JMSException: Could not create Transport. Reason: javax.jms.JMSException: Could not create Transport. Reason: java.lang.IllegalStateException: Shutdown in progress

 Enabling SSO is done via commenting out:

<!-- <authenticator class="com.atlassian.bamboo.user.authentication.BambooAuthenticator"/> -->

And enabling:

<authenticator class="com.atlassian.crowd.integration.seraph.v25.BambooAuthenticator"/>

This is done in the atlassian-bamboo/WEB-INF/classes/seraph-config.xml fle. The complete file is here.

Bamboo version in use 6.2.2.

If I remove SSO and reinstall I don't see the errors. Any hints on what might be causing this? 

[Jeremy Owen]

Hey Haani,

Unfortunately the errors included here don't point towards any root cause. They're somewhat error noise that occurs when a Bamboo shutdown has been triggered due to another failure. 

Your seraph-config.xml looks complete, there's no issues with it and my instance starts up fine with it.

Are there any other errors (ERROR, FATAL, SEVERE) further up the application log that might suggest why Bamboo can't continue to start-up when SSO is enabled?

It might be best to see if you can raise a support ticket via https://support.atlassian.com/contact/ and include an archive of your logs for our review. If you can't, let us know and we'll see what we can do to facilitate that.

Cheers,

Jeremy

[HN]

Thanks @Jeremy Owen. I've attached the full log here:
https://paste.ee/p/ucZxg

[Jeremy Owen]

Hmm. It honestly looks like something external is sending a shutdown signal to Bamboo (or the entire server is shutting down / rebooting). The server does come up and there are a few requests that do get served.

Curiously, is the server part of an Auto Scaling Group? Maybe the health check logic is failing with SSO enabled and replacing the instance. 

If this isn't the case, please raise a ticket with us at https://support.atlassian.com/contact/ and we can dig deeper.

Cheers,

Jeremy

[HN]

@Jeremy Owen thanks. I think you gave me enough to identify that an external healthcheck was causing Bamboo to restart because it did not respond within an acceptable time frame.