Description:
Systemnode starts up successfully and then goes back to stopped mode immediately.
Symptoms:
The SystemNode logs shows that the node gets another stop command from tibcohost immediately after the restart. As seen in exception below (from SystemNode logs), there is a node stop initiated by tibcohost.
***************************************
com.tibco.amx.hpa.node.Node - TIBCO-AMX-HPA-014644: Shutdown requested by TibcoHost. mode = normal, reason = node stop initiated at user request, code = -1
***************************************

This STOP command can be initiated by Administrator or tibcohost command or an external agent like Hawk and below mentioned steps helps to figure out this.

Navigate to http://localhost:8120/amxadministrator/viewaction.jsp to see if there was a STOP action created at the same time. If yes, the command came from Administrator. Since user did not actively stop the node, check if there are any pending messages in below mentioned ems queue "com.tibco.amf.admin.deploymentServerQueue.SystemHost:SystemNode". In this case we noticed one pending message in the above mentioned ems queue. The status shows "pending" and "delivered" which means the message is not acknowledged yet.

Cause:
The could happen because of some action (initiated by systemuser) which failed and became stuck in the systemnode deploymentserver queue. There was a "refresh" action for the SystemNode which failed and the message was not removed from the EMS queue "com.tibco.amf.admin.deploymentServerQueue.SystemHost:SystemNode" as it was not acknowledged. Every time systemnode is started, it is picking the pending message (refresh action) from the above mentioned deploymentServerQueue and the systemnode is going back to stopped mode.