Introduction to Replication

There are many disaster recovery solutions available today. A solution must protect against day-to-day mistakes, such as users erroneously deleting volumes, in addition to computer viruses and site disasters.

Some solutions are time-consuming and require you to back up data and manually transport the backups to a different location. Other solutions require expensive hardware and the ability to synchronously copy data across long distances, which can adversely affect application performance.

Replication technology enables you to copy volume data across groups, protecting the data from failures ranging from the destruction of a volume to a complete site disaster, with no impact on data availability or performance.

Similar to a snapshot, a replica represents the contents of a volume at a specific point in time. A replica set is the entire set of complete replicas for a volume.

A volume and its replica set are always located in different groups. By separating the groups geographically, volume data is protected against a complete site disaster.

To replicate across two PS Series groups, you must configure the groups as replication partners. The groups can be physically located in the same building or separated by some distance, depending on your needs. There must be adequate network bandwidth and full IP routing between the groups.

Each replication partner plays a specific role in the replication of a volume. The volume is located in the primary group, while the volume replicas are stored in the secondary group in space that the secondary group delegated to the primary group. Mutual authentication provides security between partners.

You can create replicas of an individual volume or all the volumes in a volume collection. When you create replicas of the volumes in a collection, the resulting set of replicas (one for each volume) is called a replica collection. In addition, you can set up schedules to automatically create replicas or replica collections on a regular basis.

The first replica of a volume is a complete transfer of the volume data from the primary group to the secondary group over the network. For subsequent replicas, only the data that changed since the previous replica is transferred to the secondary group.

If you need to transfer a large amount of volume data for the first replication, and your network link has insufficient speed or bandwidth, you can use manual transfer replication. This enables you to copy the volume data to external media and then load the data from the media to the replica set on the secondary group. After the first data transfer is complete, replication continues, as usual, over the network.

To recover volume data from replicas in the secondary group, you can clone an individual replica to create a new volume. You can also use failback functionality to temporarily host a volume from the secondary group for backup purposes or if the primary group is unavailable due to failure or maintenance. Later, you can fail back to the primary group and return to the original replication configuration, in some cases, by replicating only the volume changes that occurred while the volume was hosted on the secondary group. You can also use failback functionality to permanently switch partner roles.