IBM OS/390 User Manual

The problem management process includes the following:

•

Problem determination - the detection of the loss or impending loss of
availability of a system resource to its users, and the isolation of the
detected problem to the failing hardware, software, or microcode component.

•

Problem diagnosis - the determination of the specific cause of a problem and
the action required to resolve it. Diagnostic data gathered during problem
determination provides input to this step. It may be necessary to gather and
analyze additional information to complete problem diagnosis.

•

Problem reporting - logging or calling the appropriate group (for example, a
help desk) to have a problem logged for follow-up and solution.

•

Problem bypass and recovery - the bypass of a failure, if necessary, until a
problem can be resolved. The decision to bypass a failure is determined by
the criticality of the lost resource and the cost of providing the bypass. If
continuous operation is a requirement, recovery from a problem must take
place immediately following problem determination and diagnosis. Bypass
and recovery procedures should be automated whenever possible.

•

Problem assignment - directing the problem to the proper resolver, as
defined by enterprise policy.

•

Problem resolution - the action taken to correct a problem. Once a problem
is resolved, any steps taken to bypass it may be undone and the original
resources placed back in service

•

Problem tracking and control - tracking of problems from detection until their
final resolution. Many symptoms may result from the same problem, and
different problems may be related. Problem tracking allows the correlation of
related symptoms and problems and helps to ensure timely recovery.
Escalation of problems that exceed the established policies is a critical part
of this step.

•

Problem closing - specific notation that the problem has been solved to the
satisfaction of the reporter. Problem cause and type must be noted for
management analysis.

•

Problem analysis - analyzing problem trends to reduce the number and
impact of problems is a required management activity.

As with change management, problem management depends upon the keeping
of records about previous activities - in this case, problems that have been
reported or discovered. The problems can be reported by end users who find
anomalous behavior, system operations personnel, application programmers, or
systems programmers. Automation can also be used to report problems - for
example, if a critical job abends, NetView can detect this and create a problem
record in TME 10 Information Management, and can also monitor and update the
problem record as other detectable events occur, or at defined time intervals.

It is important that all problems be recorded in a file or database. This will
serve as a repository for all reported problems, their current status, and their
ultimate resolution, whether resolved by application change, system software
change, vendor provided maintenance, or other. As the problems are worked
toward resolution, the problem management database must be kept up-to-date.
Many smaller OS/390 installations simply use TSO/ISPF

′

s editor to create,

VSE to OS/390 Migration Workbook