Metro Mirror for ESS is built upon the functionality provided by IBM's Peer to Peer Remote Copy (PPRC or sometimes Synchronous PPRC) to provide the functions required to support Business Continuity and Disaster Recovery. Metro Mirror operates as a feature of the IBM ESS (Enterprise Storage System) and does not require any host or application code.
Metro Mirror establishes a disk to disk mirroring relationship between a primary (source) volume and a secondary (target) volume on a different physical storage subsystem. Once this relationship has been established, both volumes are updated simultaneously. This synchronous copy ensures that I/O completion is not presented to the processing system and application until the record has been successfully written to BOTH the primary and secondary volume. In this fashion, it is guaranteed that the data residing on the primary and secondary volumes are identical.
The advantage of Metro Mirror for ESS is that there is minimal host impact for performing the copy as the entire process is performed by the disk hardware. No host processing cycles or application code is used to perform the initial synchronization or subsequent updates.
The disadvantage of this type of an implementation is that since the copy operation is synchronous, there is a possible processing delays to the host application performance that increases with distance. The greater the distance between primary and secondary disk subsystems the longer it takes each I/O to complete. For this reason, Metro Mirror for ESS is usually seen only in situations where the storage subsystems are relatively close together. The often quoted rule of thumb is that Metro Mirror can be used at distances up 300 KM, or roughly 180 miles. At this distance, each synchronous write I/O would take an additional 3.6ms to complete.
The diagram at the left depicts one of the simplest Metro Mirror configurations. In this diagram there are two processors each connected to a single storage subsystem and the two storage subsystems are cabled together in order to support Metro Mirror. However, since there is no channel extension equipment depicted in this example, all of the equipment must be co-located or at least, located within physical cable distance.
A configuration like this provides for resilience and recovery in the event of a failure involving the production processor or primary disk subsystem, but will not prevent an outage in the case of a complete site failure (one involving both processors or storage subsystems) or a regional disaster that encompasses both sites.
Sometimes, a configuration like this might be used at a single site when migrating data from one storage subsystem to another.
A more resilient configuration is shown here:
This diagram features the inclusion of channel extension equipment, the necessary network resources and distance. Because of these additional components, the secondary storage and alternate processor can be located some distance away from the primary site, typically two separate physical locations perhaps several miles apart. Because of the response time penalty inherent with synchronous mirroring over distance, Metro Mirror for ESS is generally not seen at greater distances.
In this type of configuration, the enterprise can be protected from individual site failures, but is still at risk for regional disasters.
You are at the Recovery Specialties (recoveryspecialties.com) page describing the IBM "Metro Mirror for ESS" (PPRC or Peer to Peer Remote Copy) Disaster Recovery disk mirroring methodology. This is a Synchronous disk replication technique that is used for ESS (Enterprise Storage Server) that provides an exact copy of the data on a remote disk system, but at somewhat limited distance. Generally not seen at distances greater than 300km or approximately 180 miles. Metro Mirror for ESS can be used quite successfully at shorter distances - such as a campus environment without introducing too much of a distance penalty for the production application.
Recovery Specialties consulting can assist you with your Large Systems storage and Disaster Recovery needs in many ways. We are experienced in the various data replication techniques that are used today to provide Disaster Recovery and Business Continuity.
This document was printed from http://recoveryspecialties.com/