Network Deployment (Distributed operating systems), v8.0 > Migration and coexistence > Migrate Messaging resources > Migrate from WAS v5 embedded messaging > Migrate v5.1 messages using the message migration utility
Reversing the migration of messages by using the message migration utility
We can use the message migration utility to reverse migrate messaging from the default messaging provider (the service integration bus) on an application server at the current version of the product to the v5 embedded messaging provider on a v5.1 application server. Reverse migration is carried out under XA coordination.
We must reverse the migration of the messages before deleting the application server on the service integration bus.
You run the message migration utility only once for a particular queue, unless a failure occurs.
If a failure occurs during message migration, it is safe to run the message migration utility again in the same direction because messages are moved rather than copied. Whether you successfully try a failing message migration again, or delete the message, the message ordering of the remaining messages is preserved on the queue on the current version of the product.
If the reverse migration takes place after a partial migration from WAS v5.1 to a later version, in which only some of the messages were migrated, then the message ordering is not preserved. However, if the target queue in the v5.1 application server is empty when the migration takes place (as it is likely to be) then the message order is retained.
Messages that are reverse-migrated appear on the queue after those that were not migrated at the first attempt.
Procedure
- Follow the actions indicated by the message migration utility.
- On the "Select the direction of migration" panel choose Reverse a previous migration.
Results
As in normal migration, if a queue has a failing message we have the choice of:
- Trying the failing message again, in case there is a transient error.
- Delete the failing message and going on to the next message.
- Stop the queue that has the failing message and moving on to the next queue.