+

Search Tips | Advanced Search

Troubleshooting and support

There are a number of steps we can follow to help determine the nature of any problems you might encounter.

  1. Check for the following common errors:

    • The HTTP property is set to true on a route directly connected to a queue manager.
    • The SSLClient property is set to true on a route directly connected to a queue manager that is not configured to use SSL/TLS.
    • The passwords stored for the key-ring files are case-sensitive.
  2. If you find any FFST reports in the errors subdirectory, MQIPT was correctly installed but there might have been a problem with the configuration.

    Each FFST reports a problem that causes MQIPT or a route to terminate its startup process. Fix the problem that caused each FFST. Then delete the old FFST and restart or refresh MQIPT.

  3. If there is not an FFST and there is no trace output, MQIPT has not been installed correctly. Check that all the files have been put in the correct place. To check this, try to start MQIPT manually:
    1. Open a command prompt. Go to the bin subdirectory and type: 
      mqipt xxx
      where xxx is the MQIPT home directory; in this case, it is ...
    2. When MQIPT starts, look for the configuration in the home directory. Look for any error messages and FFST instances in the errors subdirectory.
    3. Look at the text output from MQIPT for any error messages. Check for instances of FFST. Correct any errors. Note: MQIPT will not start if there is a problem in the [global] section of the configuration file. A route will not start if there is a problem in the [route] section of the configuration file.
  4. If there is not an FFST but we do have trace output, configure the MQIPT connections (ConnectionLog=true) and make the sender attempt a connection. Then check that a connection from the host has been logged.

    • If a connection from the host has been logged, the sender has not been configured correctly.
    • If a connection has not been logged, check that MQIPT is configured to forward the message to the correct host and port. Then treat as a normal channel problem.