Startup Problems

The runtime instance configuration must be complete and correct to run the engine. Many problems are caused by insufficient or incorrect information in the configuration that is used by the engine.

During start-up, the runtime engine performs a number of steps that involve connections to the defined databases and transports; therefore, the configuration must be complete and correct before business processing can occur.

The following are necessary for start-up:

The directory services ASA must be running.
(The name of the service in the Windows Component Service Administrative Tool is Adaptive Server Anywhere - IO43DirectoryServices.)
The Agent must be running.
(The name of the service in the Windows Component Service Administrative Tools is Agent Launcher.)
The default port used for directory services is 7201. Verify that this port is not being used by another resource on the server.

Cannot start the instance


Error	TRANSPORT_CREATE_FAILED_MSG and [Code: 8 Msg: An MQI call failed in the WebSphere MQ Open Transport driver]
Resolution	WebSphere MQ Queue Manager and queues must be configured as defined in Unwired Orchestrator and must be running. Verify that the Queue Manager has been created and is up and running. Verify that the Queue Manager name is not misspelled in the Endpoint Manager.

Error	TRANSPORT_CREATE_FAILED_MSG and [Code: 8 Msg: An MQI call failed in the WebSphere MQ Open Transport driver] and MQI Call: MQOPEN
Resolution	Verify queues have been created in the Queue Manager Verify queue names are not misspelled in the Endpoint Manager.

Runtime database session does not start


Problem	The runtime engine fails to start and the GUI indicates that the instance is in an error state. The NNSYMessageLog.nml will contain an error message with the text "A database API call failed."
Resolution	Verify that the <runtime install dir>\instances\<instance name>\runtime directory contains the state.db and rulefmt.db files. Review the NNSYMessageLog.nml file for more detailed information. Before restarting the instance, remove the .pid file from the <runtime install dir>\instances\<instance name> directory.

Transport errors are displayed


Problem	A transport error displays in the log file.
Resolution	Verify that the queue manager or application server is running and properly configured. Verify that all physical queues are readable and writable. Using the Endpoint Manager, verify that Transport Name for each messaging service is the same name as the physical queue names.

Port conflict error messages are generated from other applications


Problem	When starting another application or component after Unwired Orchestrator has been started, port conflict errors are generated from the application or component.
Resolution	During runtime, Unwired Orchestrator Agent may use additional ports. Resolve this type of conflict by implementing one of the following: If you are running other applications or components that use ports, it is recommended that those applications or components be started before Unwired Orchestrator is started. If Unwired Orchestrator is started first, port conflicts may occur when the other applications attempt to use ports. If the first resolution option is not feasible, a more advanced resolution involves modifying the internally allocated "ephemeral" ports of the UO agent that are allocated in the Operating System TCP services. The internal allocation range used on Microsoft Windows and other operating systems may cause clashes with Sybase cooperative products if the Agent components of UO are started before them. This is manifested by the “Agent Listeners” consuming ports in the number range to which the cooperative applications are configured and will cause failure of the cooperative application when it tries to bind to port it is configured to use. To set the ephemeral (short lived) port range to the traditional configuration of the “Ephemeral Port Range”, you must configure the Operating System TCP services. The base port address for the dynamic ports is 49152. Therefore on Windows, port addresses from 1025 to 49152 are reserved, so the first dynamic port allocation begins at 49153, and sets the maximum number of ports to 65534. This is a registry adjustment. Multi-platform information regarding configuration of these settings can be at the following websites: http://www.ncftpd.com/ncftpd/doc/misc/ephemeral_ports.html http://support.microsoft.com/default.aspx?scid=kb;en-us;812873

Engine port is unavailable when the runtime instance attempts to start


Problem	When an instance is started, an error similar to the following is generated in the NNSYMessageLog.nml file: 2004-06-24 09:56:42.536\|psengine\|50624\|1\|INFR\|INFR\|Error\|?\|INFR1=67\|Failed in bind(): The socket name is already in use. 2004-06-24 09:56:42.537\|psengine\|50624\|1\|PS\|FAILED_BIND_TO_PORT\|Critical\|Engine\|Service Thread Failed to Bind to Port 7011 on host <hostname> <ip address>! No Services Available!
Resolution	Resolution requires that the port assigned for use by the engine be changed. Do the following: 1) Stop the Agent. 2) Edit the agent.properties file to change the port number range. Save the updated agent.properties file. 3) Recreate the runtime instances.

EAServer generates a message that it is starting in Administrative Mode


Problem	A message generated that EAServer is starting up in Administrative mode indicates that Unwired Orchestrator EAServer component queuing database is not started.
Resolution	If this occurs, wait a few minutes. If queues appear under Messaging in the Endpoint Manager, everything is fine. If no queues appear under Messaging, start the IOJMS database and restart IO_Jaquar.