APPC automation and recovery scenarios

(ZOS) APPC automation and recovery scenarios

This page contains scenarios and information on the automation and recovery of APPC.

Task APPC automation and recovery scenarios
Startup APPC should be started before WebSphere Application Server for z/OS . In theory, WAS for z/OS could be started before APPC, but only as long as no objects get dispatched in containers that have an IMS™ APPC LRMI associated with them. If APPC is not up before WAS for z/OS, and we want to use an APPC connector to talk to IMS, then we will have no connectivity. APPC/MVS does not have to be up for CICS . APPC does not have to be started after VTAM .
Shutdown Reverse the startup procedure. Shutdown WAS for z/OS, APPC, then VTAM.
Handling in-flight or indoubt transactions if there is a failure If we are using APPC for communications and it fails, do the following:

Shutdown all servers with APPC connectivity.
Restart APPC (if it totally failed).
Restart the WAS for z/OS server.

APPC will resynchronize itself. If our transaction is indoubt, IMS waits until you restart APPC. IMS relies on RRS for recovery. RRS will resolve transactions in doubt by handshaking with every subsystem it was communicating with before it went down. If we are using CICS, note that CICS has its own coordinator.
How to determine if APPC is running Issue the DISPLAY APPC,LU,ALL command. If APPC is not active, it will say so. In addition, the status of the logical units used by WAS for z/OS and/or IMS should be active or no APPC work will be successful.
What happens to WebSphere for z/OS if APPC goes down? Any objects attempting to use the IMS APPC PAA will not work. The server region running on behalf of the container attempting to use APPC will likely get a C9C24C05 error, indicating an APPC ALLOCATE request was attempted and failed. Additional APPC error diagnostic information that helps to pinpoint the APPC problem is contained in the logs associated with this region.
What happens to other subsystems if APPC goes down? Not applicable
Where to find more information

z/OS MVS™ Planning: Operations
z/OS MVS Planning: APPC/MVS Management
z/OS MVS Programming: Resource Recovery

Diagnose problems (using diagnosis tools)
Automation and recovery scenarios and guidelines

Task	APPC automation and recovery scenarios
Startup	APPC should be started before WebSphere Application Server for z/OS . In theory, WAS for z/OS could be started before APPC, but only as long as no objects get dispatched in containers that have an IMS™ APPC LRMI associated with them. If APPC is not up before WAS for z/OS, and we want to use an APPC connector to talk to IMS, then we will have no connectivity. APPC/MVS does not have to be up for CICS . APPC does not have to be started after VTAM .
Shutdown	Reverse the startup procedure. Shutdown WAS for z/OS, APPC, then VTAM.
Handling in-flight or indoubt transactions if there is a failure	If we are using APPC for communications and it fails, do the following: Shutdown all servers with APPC connectivity. Restart APPC (if it totally failed). Restart the WAS for z/OS server. APPC will resynchronize itself. If our transaction is indoubt, IMS waits until you restart APPC. IMS relies on RRS for recovery. RRS will resolve transactions in doubt by handshaking with every subsystem it was communicating with before it went down. If we are using CICS, note that CICS has its own coordinator.
How to determine if APPC is running	Issue the DISPLAY APPC,LU,ALL command. If APPC is not active, it will say so. In addition, the status of the logical units used by WAS for z/OS and/or IMS should be active or no APPC work will be successful.
What happens to WebSphere for z/OS if APPC goes down?	Any objects attempting to use the IMS APPC PAA will not work. The server region running on behalf of the container attempting to use APPC will likely get a C9C24C05 error, indicating an APPC ALLOCATE request was attempted and failed. Additional APPC error diagnostic information that helps to pinpoint the APPC problem is contained in the logs associated with this region.
What happens to other subsystems if APPC goes down?	Not applicable
Where to find more information	z/OS MVS™ Planning: Operations z/OS MVS Planning: APPC/MVS Management z/OS MVS Programming: Resource Recovery