Issue link: https://epubs.iltanet.org/i/4636
ILTA White Paper Infrastructure Technologies 12 happen prior to replication, with the resulting data compressed and optimized by the WAN accelerators. While straightforward, the issue with using SnapMirror technology alone is its inability to address open files, database consistency and powered-on VMs — similar to rsync and Microsoft Distributed File System (DFS) in the original environment. While posing no problems for CIFS files, the result with a VM is a crash-consistent backup, where the backup state would be equivalent to an unexpected power-off. BACKUP NetApp's SMVI fully integrates with VMware's vCenter server to create snapshots of virtual machines utilizing VMTools and Microsoft VSS as with vRanger. This creates a complete backup of the server and data files. The integration of SMVI with SnapMirror allows it to trigger a replication of data to the DR SAN once the VM snapshots are completed. Like vRanger, SMVI allows for the restoration of the most recent VM backup and also creates incremental backups. The next version promises to make file and volume restoration available. The benefit of this technology is that the work is done at the SAN level. This eliminates the need for a backup agent, other than VMTools, which is otherwise required, in each guest OS. This takes out any CPU load and network traffic that would otherwise be required by the guest OS during backup. While this conceptually eliminates both the disk-to- tape and disk-to-disk backup of each individual VM, we continue to use i365's EVault for backup of core systems for long-term data retention. DISASTER RECOVERY With the VMs and data at the DR site, the final piece of the puzzle is a solution to make the necessary changes — such as updating the IP address on the replicated VMs — so that they function at the DR site and can be put into production. For this we use VMware's Site Recover Manager (SRM), which, like NetApp's SMVI, fully integrates with VMware and NetApp to provide a comprehensive solution. SRM provides a single interface for execution of the DR recovery process. SRM allows for the primary site VM shutdown, DR site VM start-up and guest OS network changes. It can also include user- defined prompts and scripts for required tasks that are outside of the VM infrastructure. This feature has allowed Davis to create recovery plans tailored to various levels of disaster scenarios and user sophistication, to allow for both a sophisticated IT member and an average user to initiate the recovery process. The latter is in place to ensure the DR process can be initiated even without available IT resources. SRM also has the ability to export the recovery plan steps, which documents the entire recovery process, "With the VMs and data at the DR site, the final piece of the puzzle is a solution to make the necessary changes . . . so that they function at the DR site and can be put into production."