If you can reproduce the problem, consider
the conditions under which you can do so:
- Is the problem caused by a particular message flow? If so, use the debugger facility
of the workbench and user
tracing to identify the problem.
- Is the problem caused by a command? On distributed
operating systems, you issue commands at the system command line.
On z/OS, you can issue commands from the console,
the syslog, or by submitting a batch job. You enter customization commands
from an OMVS session. Console commands that you enter from the console or
syslog might be converted to uppercase, depending on the system configuration.
This can cause some commands, such as mqsichangetrace,
to fail, especially if these commands contain parameters that must be lowercase.
An error message indicating that the execution group is not available might
be caused by the execution group name being in the wrong case. The same thing
can happen on message flows.
- Does a problem command work if it is entered by another user
ID?
If the command works when it is entered by another
user ID, check the environment of each user. Paths, especially shared library
paths, might be different. On Windows, UNIX systems, and Linux verify that all
users have set up their command environment correctly; refer to sample
profile for more information.
On Windows,
the environment for the broker is determined by the system settings, not by
a particular user's variables. However, the user's variables affect non-broker
commands.
On LinuxUNIX systems, only the service ID that is
specified when the broker was created can start a broker.
On Windows, any authorized user can start a
broker.