Anatomy of a migration

As depicted in Figure 1, the migration of an IBM i is broken down into three stages.

1. System Migration – Creation of the copy node

   During the system migration stage, Migrate While Active is configured and a system save is performed on the source node. The save includes the operating system and some or all of the existing user data. The system save is done while in restricted state. The Migrate While Active user indicates how much user data to save. All user data saves are focused on objects within SYSBAS.

   A unique version of install media is created during the system save and used to scratch install the copy node. Then any saved user data is restored onto the copy node.

2. Data Synchronization – Keep the copy node "in sync" with the source node

   After the system save of the source node is complete, users and applications can run production workloads on the source node. During this time, objects within SYSBAS libraries and the integrated file system can be created, changed, or deleted. Any change to these objects will be noticed by Migrate While Active and tracked.

   After the system migration is complete, Migrate While Active begins the continuous replication stage that is known as data synchronization. During this time, Migrate While Active is continuously but asynchronously keeping the copy node "in sync" with the source node.

3. Cutover – Complete the migration

   The Migrate While Active user decides when to leave the data synchronization stage and finalize the migration to the copy node.

   During the cutover stage, the source node enters restricted state and a final synchronization of any remaining tracked objects occurs. After the final synchronization is complete, the Migrate While Active user decides whether to complete the migration. Upon successful completion of cutover, data synchronization between the source and copy nodes is ended and the copy node is ready to be used.