You can use SyncMirror to mirror aggregates, and thus provide increased data resiliency. SyncMirror removes single points of failure in connecting to disks or array LUNs.
What SyncMirror is
SyncMirror is an optional feature of Data ONTAP. It is used to mirror data to two separate aggregates. It allows for real-time mirroring of data to matching aggregates physically connected to the same storage system.
Advantages of using SyncMirror
A SyncMirror aggregate has two plexes. This setup provides a high level of data availability because the two plexes are physically separated.
What mirrored aggregates are
A mirrored aggregate is a single WAFL storage file system with two physically separated and synchronously up-to-date copies on disks or array LUNs. These copies are called plexes. Data ONTAP typically names the first plex plex0 and the second plex plex1.
The states of a plex
A plex can either be in an online state or in an offline state. In the online state, the plex is available for read or write access and the contents of the plex are current. In an offline state, the plex is not accessible for read or write.
Splitting a mirrored aggregate
Splitting a mirrored aggregate removes the relationship between its two plexes and creates two independent unmirrored aggregates. After splitting, both the aggregates come online.
Rejoining split aggregates
You can rejoin split aggregates. You might want to do this if you have set up an HA pair in a MetroCluster configuration and a disaster breaks the HA pair.
Removing a plex from a mirrored aggregate
You can remove a plex from a mirrored aggregate. You might do this if you want to stop mirroring the aggregate, or if there is a problem with the plex. Removing a plex results in an unmirrored aggregate.
Comparing plexes of a mirrored aggregate
The plexes of a mirrored aggregate are automatically synchronized. You can confirm this by using the aggr verify command to check for, and in rare cases, fix any inconsistencies.
How to avoid unnecessary RAID reconstruction when a plex fails
If a plex in a mirrored aggregate fails due to fabric failure such as a cable disconnection or a Fiber Channel switch power failure, you can manually take the failed plex offline before the cable connection or power supply is restored to avoid unnecessary RAID reconstruction.