Virtualization is booming. But as increasing numbers of businesses incorporate the technology, it’s important to remember that consolidating servers onto fewer physical machines comes with certain risks. Yes, businesses of any size can realize terrific resource benefits, but the cost of server failure can be high. A given host server can become quite valuable as it holds more and more virtual machines (VMs): If it goes down or experiences problems, business operations can be severely impacted. For this reason, you need a potent high availability/disaster recovery solution.
Microsoft and VMware offer high such functionality within their industry-standard products, and those features provide good, basic protection. However, for ease of use or more granular functionality, you should consider third-party products such as those covered in this buyer’s guide.
What Virtualization Adds
To restore a physical environment, you have to laboriously re-install the OS and all applications; in a virtual environment, you’re dealing with hardware-independent VMs—mere collections of files— that you can simply copy to alternative locations. Starting up the backed-up VM is a matter of dragging and dropping it to a newly operational host and starting it up.
Of course, virtualization in itself doesn’t give you an automatic high availability/disaster recovery safety net. You still need an effective disaster-recovery plan for your VMs—just as effective a plan as you would have for physical machines. But virtualization does, in fact, greatly simplify disaster recovery, particularly if you can—for example—utilize one of this buyer’s guide’s products to replicate your running VMs offsite, where they can await activation.
These products automatically transfer and restart your VMs at a predefined location; they can even start VMs in a certaind order to proactively resolve any dependencies associated with the VM.
A prime method of ensuring high availability in a physical environment is failover clustering—and it’s important in virtual environments, too. Failover clustering does a great job of increasing the availability of VMs—as well as the applications hosted inside those VMs—on the occasions of unplanned (and even planned) downtime. When the host server goes down, a standing-by host can automatically take its place, with working VMs ready to go. Microsoft’s Hyper-V and VMware’s VMware Server let you manage such cluster scenarios—for example, you can use VMotion or Live Migration to move and monitor workloads hosted in VMs.
Third-party tools can extend the protection offered by failover clustering across geographical boundaries, providing automated failover and failback. Primarily, these tools can provide geographical data protection, and of course they’re just plain easier than implementing multisite clustering.
Considerations While Shopping
There are many considerations to keep in mind while checking out the market. Depending on your existing environment, the replication method is crucial. Will you use a SAN-based snapshot? Or will you use an image/snapshot of the VM directly? Another question to ask: Are all OSs treated equally? Will a Windows-based VM have an online snapshot backup taken, but a Solaris-based VM have to have a file-level or offline backup taken? What about support for mixed virtualization environments? You might have only VMware ESXi now, but you’d like to be prepared next year, when one of your users needs to use Hyper-V.
Third-party tools simply add functionality to the basic feature sets available from Microsoft and VMware—or at least extend existing functionality. And in some cases, third-party tools might offer the same features but just do things better! Finally, watch for tools that ofer smooth integration with suites or tool collections that you already have from the same vendor. If you’re familiar with one company’s tools, that familiarity will come in handy when using new tools.
A great product will provide that kind of entegration with your existing infrastructure and capabilities to provide a comprehensive high availability/disaster recovery solution. As virtualization becomes more and more popular in the enterprise, high availability/disaster recovery capabilities are vital to business continuity. Remember, disasters can come from many directions: Natural disasters, power loss, and simple human error can wreak equal havoc on your systems. So, hopefully, you’re thoroughly testing your plan to make sure it works fluidly and effectively—and that advice holds true whether your environment is real or imaginary.