1154 - How to employ failover and failback measures in migration

Migrating the AccountServer

When an AccountServer needs to be migrated to new hardware or a new site, or when it is decommissioned, failover and failback measures can be employed to ensure that no data loss occurs. Simply follow the steps below.

1. Failover (see Article 1152) to the Slave AccountServer.
2. Move the original (Master) AccountServer to its new location as required.
3. If the Master is to be decommissioned, create a new Slave AccountServer, and fail back to it.
4. If the Master is to remain in use, fail back to it.

Migrating the StorageServer

When a StorageServer needs to be migrated to a new data centre or host, you may prefer to do this by promoting a MirrorServer to a StorageServer. Failover and failback measures can be employed to ensure that no data loss occurs. Simply follow the steps below.

1. Promote the MirrorServer to a StorageServer and failover to this StorageServer (see Article 1152).
2. Move the original StorageServer to its new location as required and turn it back on. Alternatively, if the original StorageServer is being decommissioned, you can create a new MirrorServer to fail back to.
3. Run Integrity Checks on the data on both the original StorageServer and on the promoted MirrorServer.
4. Mirror back to the original StorageServer, and run an Integrity Check again.