Metro provides bi-directional synchronous replication (active/active) across two
PowerStore systems. A metro volume is exposed using two distinct systems, typically in two different data centers, up to 96 km (or 60 miles) apart, or in two distant locations within the same data center. The two systems cooperate to expose a single metro volume to application hosts by providing the same SCSI image and data. The hosts and application perceive the two physical volumes that are hosted by the two systems as a single volume with multiple paths.
Metro protection enables increased availability and disaster avoidance, resource balancing across data centers, and storage migration between two
PowerStore systems.
When you configure a metro volume, the content of the volume is replicated to the remote system. Protection policies are used to configure additional protection such as local snapshots.
A metro session consists of two
PowerStore systems and optionally a witness server.
When you configure a metro resource, the system from which the metro source is configured is automatically set as preferred and the other is configured as nonpreferred. When no witness is configured or when the witness is unavailable, these roles help to guide the system behavior on failure situations. When a failure occurs (either on one of the systems or to the connection between the systems), the metro session becomes 'fractured', and the nonpreferred system stops servicing I/Os while the preferred system provides host access.
The witness server is a passive third party that is installed on a stand-alone host (preferably in another data center so it is not affected by power failures to the
PowerStore systems). The witness observes the status of the two systems. When failure occurs, the witness server determines which system remains accessible to hosts and continues to service I/Os. A witness that is installed on a third site provides protection from single failure scenarios.
Metro switches between using the witness and using the system role as means for recovering from single failure situations (when the witness is not configured or is unavailable, recovery from a single failure is done manually).
For a summary of metro attributes and comparison to synchronous and asynchronous replication, see
Replication summary.
Data is not available for the Topic
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please select whether the article was helpful or not.
Comments cannot contain these special characters: <>()\