Unserviceable routing requests
If you operate BTS in a sysplex, you can route processes and activities across a set of CICS® regions called a BTS-set. When a process or activity is started by a RUN ASYNCHRONOUS command, it can be routed either statically or dynamically. Mostly, you might choose dynamic rather than static routing.
For detailed information about routing processes and activities, see Administering BTS.
Static routing
Using static routing, you name the target region to which the activity is to be routed on the REMOTESYSTEM option of the installed transaction definition, for the transaction associated with the activity. If the target region is unavailable at the time the activity is to be started, CICS treats the request as unserviceable.
Dynamic routing
Using dynamic routing, the target region is chosen by your routing program, the distributed routing program or the CICSPlex®Â SM routing program. If the target region that it returns is unavailable, the routing program is invoked again and can select a different target. Alternatively, it can (by setting a non-zero return code) indicate that the request is to be treated as unserviceable.
For definitive information about writing a distributed routing program, see Writing a distributed routing program.
Why classify requests as unserviceable?
Why should your routing program classify requests as "unserviceable"? Why should it not reroute the request to an alternative region, assuming that alternatives are available?
Sometimes, due to a transaction affinity, it might be essential that an activation runs on a specific region, and no other. If so, rather than selecting an alternative target region, your routing program can return the same target (even though it is currently unavailable), and classify the request as unserviceable.
How CICS handles unserviceable requests
CICS identifies unserviceable requests by issuing a DFHSH message and tries the request again every minute.
- Issues message DFHSH0105, which identifies the request and indicates that it cannot be serviced.
- Tries the request again every minute. If the request is successfully serviced, CICS issues message DFHSH0108.
- Each hour, if the request still cannot be serviced, issues message DFHSH0106. This message indicates the time remaining before CICS purges the request, if it has not been serviced in the meantime.
- After 24 hours, if the request still cannot be serviced, stops trying to service it and issues message DFHSH0107. The request is discarded.
Resolving unserviceable requests
In many cases, CICS resolves unserviceable requests automatically. If, for example, an unavailable target region becomes available within 24 hours of the request being issued, CICS routes the request correctly.
You should watch for occurrences of DFHSH0105 and DFHSH0106 messages. You should investigate why the request is unserviceable, and take any necessary corrective action. It might be, for example, that a resource required to satisfy the request (an activity or process) is inaccessible; or that a remote region, or a link to it, is unavailable.