Scenario: You issue the mqsistop command to stop the broker, but the system
freezes and never actually stops any of the execution groups.
Explanation: One possible cause is that you or someone else
is debugging a flow and it is currently stopped at a breakpoint. WebSphere Business Integration Message Broker regards
this as a message in flight situation, and refuses to stop the broker through
the normal command.
Solution: Click Stop
debugging from the workbench (or
ask the person doing the debugging to do this). After that operation has completed,
the broker stops.
If is not possible to stop debugging, end
all execution group processes associated with that broker to allow the broker
to stop. However, your messages are backed-out and you still have to click Stop debugging after the broker restarts.
You cannot stop the broker queue manager
Scenario: You are trying to use the WebSphere MQendmqm command to stop a broker queue manager
on a distributed platform, but it doesn't stop.
Explanation: In certain circumstances, attempting to stop
a broker queue manager does not cause the queue manager to stop. This can
happen if you have configured any message flows with multiple threads (you
have set the message flow property Additional Instances to a number greater
than zero).
Solution: If you want to stop the broker's queue manager,
stop the broker using the mqsistop command
specifying the -q flag. (The -q flag is not available on z/OS.) This
issues the WebSphere MQendmqm command
on your behalf in a controlled fashion that shuts down the broker and the
queue manager cleanly.
Execution group terminates abnormally
Scenario: Your execution group processes terminate abnormally.
Explanation: When execution group processes terminate abnormally,
they are automatically restarted by the bipbroker process. If an execution
group process fails, it is restarted three times during each five-minute interval.
The first five-minute interval begins when the execution group is first started.
The broker database table BROKERAAEG contains a column labelled RetryInterval.
This column defines the restart time in minutes. Each row in this table represents
an execution group. The row is populated on the first deploy of an execution
group and RetryInterval defaults to 5.
If you remove the execution
group from the broker configuration, deploy the broker configuration, and
then later add the execution group and redeploy the broker configuration.
The row is re-created and RetryInterval is set to its default value of 5.
Solution: To change the default value:
Stop the broker.
Change the RetryInterval's value in the database table.