Having a virtualization disaster recovery plan is a great way to protect small and medium-sized businesses' most critical data. Backups simply won't cut it.
Backups are of little help when a disaster wipes out a data center. More importantly, data is worth little without servers and applications. That's where a virtualization disaster recovery plan comes into play.
Virtual disks are far easier to replicate to alternative sites because of the ways they are backed up. In addition, many disk-based backup tools offer off-site replication for data. What results is an affordable virtualization disaster recovery strategy that's baked into your backup solution.
Key ingredients for a virtualization disaster recovery plan
Integrating backups, replication and disaster recovery is a perfect fit for the needs and budget of a small and medium-sized business (SMB). Here's what you'll need to get started:
A disk-based backup tool. Look for a product that supports off-site replication. The right one will eliminate backup windows as well. Instead of a backup every night, you now will have a constant and steady stream of changed virtual disk blocks on disk. You'll want a tool that accomplishes this with little effect on resource use on your virtual machines (VMs) and hosts.
An alternate site. Previously, one of disaster recovery's biggest expenses was the cost to maintain a completely separate "hot site" that's immediately primed for failover. But an SMB's recovery-time objective (RTO) after a disaster can be measured in days, instead of minutes or seconds. If this RTO describes your organization, replicating your most-critical VMs to a cloud storage provider or to alternate disks in a colocation facility might be good (and cost-effective) enough.
A network connection. Replication requires a network connection. For most SMBs, a reasonably speedy Internet connection is required for off-site replication. Your mileage will vary, but today's SMB-friendly backup solutions that focus on disk blocks offer the best solution for conserving the amount of bandwidth you'll actually need. Estimating the exact amount is a bit of an art form, but some backup tools provide tools that'll deliver a best guess.
Disks and servers. You'll need servers and disks at your alternate site for storing the virtual disks. Depending on your needs, you can save some money here. You don't necessarily need high-speed disks, unless you intend to restart VMs from the secondary site after a disaster.
As an alternative, you could negotiate with your hardware provider for a deliver-on-disaster arrangement. This approach allows you to use low-speed disks for daily replication and bring in additional horsepower only when it's necessary for disaster operations. This approach stands in contrast with that of a cloud provider, which can spin up VMs almost immediately when disaster strikes, and charges you only for what you use.
Implementing a virtualization disaster recovery plan
With all the physical pieces in place, implementing a disaster recovery plan actually can be quite simple. With the right backup solution, turning on replication could be as easy as making a few clicks in the administrator console.
More on creating a virtualization disaster recovery plan
Planning for Hyper-V disaster recovery on the cheap
From there, give the backup tool the network address for its off-site partner, and keep an eye on your bandwidth utilization. Remember, replication always takes time at first. So, you'll see the savings once the initial virtual-disk images finish transferring.
Be aware that not every disaster is a force of nature. Tornadoes, hurricanes, earthquakes and other natural disasters come to mind easily, but these catastrophes rarely occur. The disasters that are far more common include corruption of a server database, equipment failure or OS issues that stem from a bad patch. These are nowhere near as painful as classic natural disasters, but they still will affect users greatly.
Other backup features to look for are built-in failover and failback. Good tools can power-on the alternate VM, quickly getting that machine back to servicing users. With this ability, you have the freedom to fail over workloads when a problem occurs. You can reverse the process later and move the replication of data back to the original server (or to a fresh replacement) once you resolve the problem.
Once a viable option only for IT's rich and famous, truly affordable virtualization disaster recovery is now an almost simple implementation for SMBs. Virtualization absolutely is a game changer, but be aware that you shouldn't focus exclusively on your VMs. Some backup solutions can even back up and replicate your physical servers as well.
What's even more heartening is that backup vendors have been working for a number of years on products that support these features. Virtualization backup products are mature, and the options are vast. Arguably, the hardest part is choosing which backup solution for disaster recovery is the right one.