Typically this happens because the
customer has pointed the SMP/E SBBOLIB DDDEF statement to a service HFS
(such as /service/WebSphere390) and failed to copy the service HFS to the
production HFS after applying service.
Another possible cause is that the customer did
not remove old copies of load modules from the LPA or did not refresh
LLA.
This problem is extremely difficult to pinpoint
because the symptoms will vary depending on the from and to service levels
and other factors. The primary clue is that the problem occurs when the
Daemon or server address spaces are initialized, during an IVP run or when
a new administrative conversation is activated soon after service is
applied.
In order to confirm the diagnosis, the customer
must validate that the LPA and LLA have been rebuilt as required and that
any service HFS files have been successfully copied to production
locations.
Individual occurrences of the problem are solved
by copying the required service HFS to the production location or
refreshing and rebuilding the LLA and LPA as required. Since most
customers who employ a service HFS do so for good reason, changing the
DDDEF statements to point to the production HFS often is not
practical.
The customer can only prevent this problem by
establishing rigorous change control procedures and checking for the
success of his HFS copy jobs.
|