In a volume migration operation, the data on the source volume is copied to a target volume. After all data is copied to the target volume, host access is transferred to the target volume to complete the migration. During the migration, the source volume remains available for read and write operations.
The volume migration operation copies the content of the source volume to the target volume. If the source volume is updated by write operations during the copy operation, the updates will be recorded on the differential table of a volume. If there are differential data created by the update, the differential data are copied from the source volume to the target volume. This process is repeated until all differential data on the source volume is copied.
The following figure shows the data flow during a Volume Migration operation.
There is an upper limit to the number of times a copy operation is executed, and the limit depends on the capacity of the source volume. The limit increases as the capacity of the source volume increases. When differential data still exists after copy operations are repeated to the upper limit, the migration fails. Before attempting the migration again, the workload between the host and the storage system must be adequately reduced, preferably to a value below 50 IOPS (input/output operations per second).
After the volumes are fully synchronized (that is, there is no differential data on the source volume), the storage system completes the migration by swapping the Reserve attribute (from target volume to source volume) and redirecting host access to the target volume. The status of the target volume becomes normal. Volume Migration performs all migration operations synchronously (that is, one migration at a time). The following figure illustrates the state of the volumes after a Volume Migration operation.