.:: a few thoughts on the subject by rob wright ::.

I’m new to WebSphere MQ and have been struggling to learn the best way to debug problems. Of particular confusion has been the term “logs.” If you create a persistent queue then every message is persisted in what MQ calls a log. This is dramatically different from the way I usually use the term. When I say log, I usually mean a file that contains trace or debug information. I would call the MQ log a database.

So everytime I search the manual for “log information,” I get nothing that helps me debug a problem. Today however, I did run across a gem of a command, mqrc. As with many legacy applications, MQ displays error or reason codes when an error occurs. So you get a very descriptive “2051″ to help you determine the problem. For a given reason code, mqrc displays the code definition…a real description behind the number. Below is the help for that command.

MQRC Usage:

Interpret a retcode or a number of retcodes in a range
mqrc [-a]
mqrc -r [-a]
mqrc -R [-f -l ] [-a]

Interpret a message or a number of messages in a range
mqrc [-a] AMQ
mqrc -m [AMQ] [-a]
mqrc -M -f -l [-a]

Interpret a retcode symbol
mqrc
mqrc -s

Display help
mqrc -h Display this help
mqrc -v Display the version number

Where the options are:

-a Try all severities to find message text
-f First number in range
-l Last in number range

1. Numeric arguments are interpreted as decimal if they start with a
digit 1 to 9, or hex if prefixed with “0x”

2. If no range is given when interpreting a range of retcodes, then all
the retcodes are listed.

3. When printing retcodes or messages in a range, if there is a problem
with message, an indication is printed just before the message text,
as follows:
‘?’ There are no matching retcodes for this message
‘!’ The message severity is different to the retcode severity