NetApp gives MetroCluster a good going-over
From minutes to seconds
NetApp has updated its MetroCluster software to support storage array failover to remote site storage in VMware environments.
MetroCluster is software that, using replication, synchronously mirrors data writes from a primary NetApp FAS array to one in a remote site, across a campus or metro area, up to 100km away. It runs in the NetApp array controller and will failover user access to the remote site if a component in the array, below the controller, fails. The NetApp arrays in the primary and secondary sites form a high-availability cluster and are seen as a single resource by accessing servers.
NetApp says that, as a result of integration testing with VMware, the MetroCluster product now offers continuous data availability in VMware environments with failover accomplished in seconds. No update to NetApp's Data ONTAP operating system is involved.
It also facilitates transparent (to the users) software and hardware upgrades of a NetApp storage array with user data accesses re-directed to the remote site during the upgrade process and then returned to the primary array. Additionally the MetroCluster software has been integrated with NetApp's ASIS deduplication to reduce the amount of stored data.
The software works with VMware's ESX server and any storage array failover is transparent to ESX except in the situation where the array controller itself fails. In that case it cannot carry out its failover duties which have to be performed manually.
NetApp's SnapMirror software can be used to extend data availability across distances longer than 100km by using asynchronous mirroring. ®