Application outages may occur for any number of reasons. Whether hardware failures, software bugs, user error, or a ransomware attack, these events cause application downtime. Since application downtime has multiple deleterious effects on organizations, fast application recoveries have taken on new importance. In response, virtual machine (VM) backup software providers have introduced and implemented multiple new instant recovery features into their solutions.
Instant Recovery
Instant recovery provides organizations with the opportunity to recover applications instantly without going through normal restoration processes. All backup solutions now utilize disk as a backup target. As a result, these backup vaults residing on disk have become viable options to use for hosting product application data.
Further, most organizations host most if not all their applications and data on virtual machines (VMs) running on hypervisors. Enterprise-grade hypervisors offer native software tools to dynamically migrate these VMs from one storage platform to another. These tools position organizations to initiate the recovery of a production VM on backup disk and then migrate it to production.
Innovations
This option to perform instant recoveries from backup data has resulted in multiple innovations from backup solution providers. While most enterprise backup providers support instant recovery, each one offers different techniques to deliver it.
Here are a few examples of how some providers deliver instant recoveries in their backup solutions:
Cobalt Iron Compass
Cobalt Iron offers two instant recovery options.
Instant Access. Using this feature, enterprises may access, run, and modify a VM without restoring the data. Compass performs this instant access restore while maintaining an immutable stored version of the VM’s data.
In this restore type, Compass creates a definition of the VM in vSphere without restoring its data. Once started, the VM runs directly against the Compass backup repository. When shut down, Compass removes the VM from vSphere along with any resources it used. Enterprises may use these restores for analytics, development, or any use cases that only require access to a VM’s data.
Near-instant Restores for very large VMs. Enterprises may perform near-instant restores for up to multi-terabyte VMs through vSphere using Compass’ vSphere client plug-in. Compass instantly restores a VM with full read-write capabilities when it restores it.
During the restore, administrators may create a new VM or replace the existing one. To facilitate rapid, non-disruptive restores, administrators choose the disks to recover and the mount proxy to use. Selecting between available mount proxies becomes critical if restoring multiple VMs as this helps minimize contention during recoveries.
Unitrends Unified BCDR
VM Replicas. Unitrends’ instant recovery VM replica feature ensures enterprises may instantly bring up a VM should a production VM fail. Its offering differs in a slight but significant way from the instant recovery feature other solutions offer.
Unitrends uses the most recent backup of the original VM as the source of the data for the replica. It then creates the replica VM on a specified ESXi host. Unitrends configures the replica as a cold standby so it consumes no compute resources. Unitrends then keeps it current by queuing a replica restore job each time a backup successfully completes. In this way, if and when an organization wants to bring up that replica, it is instantly ready to go on an ESXi host.
Veritas NetBackup
Instant Rollback. Introduced in NetBackup 9.1, Instant Rollback equips administrators to select one or multiple VMs to simultaneously recover to a previous backup. NetBackup then restores the selected VM or VMs into production by only restoring the changed blocks.
It differs from other instant recovery offerings as it performs instant recoveries at scale. Administrators may simultaneously schedule restores for hundreds or even thousands of VMs. If needing more than 100 restores, NetBackup schedules the request as multiple batch jobs. Each batch job then performs up to 100 VM restores.
Just a Sample
Access to instant recovery is a feature many backup solutions natively offer. This should help put to rest concerns organizations have about their ability to recover quickly.
However, the examples from these three solutions represent only a few ways backup software solutions now deliver instant recoveries. Some host instant recoveries on the appliances they offer. Some deliver instant recoveries in general-purpose, purpose-built or both types of cloud environments. Some start recoveries on their appliances even as they dynamically migrate data back into production in the background.
These listed examples do, however, offer insight into the new instant recovery options to which organizations have access. Some implementations excel at instantly recovering and running VMs in production. Others excel at scheduling the recovery of lots of VMs but which may take some time to recover fully. Still others excel at granting instant access to the application and data but not running the application in production mode.
Taken together, organizations should NOT assume all backup solutions implement instant recovery in the same way. Rather, organizations should assume they implement instant recovery differently. These different implementations may and should influence the choice an organization makes for its VM data protection solution.
Keep Up to Date with DCIG
To be notified of new DCIG articles, reports, and webinars sign up for DCIG’s free weekly Newsletter.
Technology providers interested in licensing DCIG TOP 5 reports or having DCIG produce custom reports, please contact DCIG for more information.