Home

 

Starting a backup queue manager

 

To substitute an unrecoverable queue manager with it's backup queue manager, do the following:

1. Execute the following control command to activate the backup queue manager:

   strmqm -a BackupQMName

   The backup queue manager is activated. Now active, the backup queue manager can no longer be updated.

2. Execute the following control command to start the backup queue manager:

   strmqm BackupQMName

   WebSphere MQ regards this as restart recovery, and utilizes the log from the backup queue manager. During the last update to the backup queue manager replay will have occurred, therefore only the active transactions from the last recorded checkpoint are rolled back.

   When an unrecoverable queue manager is substituted for a backup queue manager some of the queue manager data from the unrecoverable queue manager can be lost. The amount of lost data is dependent on how recently the backup queue manager was last updated. The more recently the last update, the less queue manager data loss.

3. Restart all channels.

Check the resulting directory structure to ensure that you have all the required directories.

See Directory structure (Windows systems) and Directory structure (UNIX systems) for more information about WebSphere MQ directories and subdirectories.

Make sure that you have a log control file as well as the log files. Also check that the WebSphere MQ and queue manager configuration files are consistent so that WebSphere MQ can look in the correct places for the restored data.

If the data was backed up and restored correctly, the queue manager will now start.

Even though the queue manager data and log files are held in different directories, back up and restore the directories at the same time. If the queue manager data and log files have different ages, the queue manager is not in a valid state and will probably not start. If it does start, your data is likely to be corrupt.

 

Parent topic:

Backing up and restoring WebSphere MQ

fa14750_

 

Home