An unserviceable request is a request that cannot currently be satisfied. It may be, for example, that an activity is not available, or that the region on which the request is to execute is not accessible.
If you operate BTS in a sysplex, it is possible to route processes and activities across a set of CICS® regions called a BTS-set. (For detailed information about routing processes and activities, see Using BTS in a sysplex. )
When a process or activity is started by a RUN ASYNCHRONOUS command, it may be routed either statically or dynamically. Mostly, you will probably choose dynamic rather than 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.
Using dynamic routing, the target region is chosen by your routing program (the distributed routing program or the CICSPlex® SM routing program). If the target region that it returns is unavailable, the routing program is reinvoked 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 the CICS Customization Guide.
Why should your routing program classify requests as "unserviceable"? Why should it not simply re-route the request to an alternative region, assuming that alternatives are available?
Sometimes, perhaps due to a transaction affinity, it may be essential that an activation should run on a specific region, and no other. If this is the case, 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.
When a request is "unserviceable", CICS:
In many cases, CICS will resolve 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.
Your operators should watch for occurrences of DFHSH0105 and DFHSH0106 messages. They should investigate why the request is unserviceable, and take any necessary corrective action. It may 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.