Understanding nonswitched MetroClusters

What nonswitched MetroClusters do

Like mirrored clusters, nonswitched MetroClusters contain two complete copies of the specified data volumes or file systems that you indicated as being mirrored volumes or file systems in your cluster. These copies are called plexes and are continually and synchronously updated every time Data ONTAP writes data to the disks. Plexes are physically separated from each other across different groupings of disks.

Unlike mirrored clusters, MetroCluster provides the capability to force a failover when an entire node is destroyed or unavailable.

Configuration overview

The MetroCluster configuration employs SyncMirror and a cluster configured for mirrored volumes to build a system that provides data even after complete loss of one of the cluster nodes and the storage at that site. All mirrored volumes (both root volumes and data volumes) must be mirrored with one plex at each site to preserve data using MetroCluster.

Attention
The MetroCluster configuration can preserve data only if volumes are mirrored. Unmirrored volumes are lost if the storage where they reside is destroyed.

See the Data Protection Online Backup and Recovery Guide for detailed information about using SyncMirror to mirror data.

The following figure illustrates the MetroCluster configuration.

Note
DS14mk2 FC disk shelves using ESH or ESH2 modules, or DS14mk2 AT disk shelves using AT-FCX modules operating at 2-Gb Fibre Channel speeds, can attain a maximum node separation distance of only 300 meters, not 500.

Advantages of nonswitched MetroClusters

MetroClusters provide the same advantages of mirroring your data as standard mirrored clusters do:

Your data is protected in the event of a failure or loss of two or more disks in a RAID4 aggregate or three or more disks in a RAID-DP aggregate.
The failure of an FC-AL adapter, loop, or ESH module does not require a cluster failover.

In addition, a MetroCluster enables you to initiate a failover if an entire site becomes lost or unavailable. If a disaster occurs at one of the node locations and destroys your data there, your data not only survives on the other node, but can be served by that node while you address the issue or rebuild the cluster.

For more information about recovering from a disaster using MetroCluster, see Disaster Recovery Using MetroCluster.

How Data ONTAP works with nonswitched MetroClusters

During configuration of the cluster, Data ONTAP identifies spare disks and divides them into separate groupings called pools. These pools of disks are physically separated from each other, allowing for high availability of mirrored volumes. When you add a mirrored volume or add disks to one side of a mirrored volume, Data ONTAP determines how much storage you need for the second half of the mirror, and dedicates that storage from a separate pool to the mirrored volume.

Setup requirements and restrictions

The following list specifies the requirements and restrictions you must be aware of when setting up a nonswitched MetroCluster.

Architecture compatibility: both nodes must be the same type and model.
Disks in the same plex must be from the same pool, with disks in the opposite plex from the opposite pool.
The SyncMirror pool rules for direct-attached Fibre Channel disks must be observed.
For more information, see SyncMirror pool rules for direct-attached Fibre Channel disks.
There must be sufficient spares in each pool to account for a disk failure.
The following licenses must be enabled on the local and partner nodes:
- cluster
- syncmirror_local
- cluster_remote
  Note
  See the "Clustered Failover with 500 Meter Distance" document on the NOW^TM site for more information about hardware requirements for this configuration.