Virtual Observer

No More Bad Luck with Properly Backed Up VMs

If you have a virtualized backup and disaster recovery plan, you don't need to count on luck for your business to remain operational any day of the year.

For the superstitious, Friday the 13th is a day of bad luck, when everything goes wrong. I'm not superstitious, and when it comes to protecting my VMs I'm not worried. I've taken the proper precautions. In the modern data centers, there is no room for error. Expectations today are very high, indeed.

That said, I'd bet there are a lot of folks out there who worry a lot about luck because they're not prepared for the worst. Say, for instance, the water main up the street from your data center bursts and turns your data center into a small pond? Or a tornado wipes it out? Or an earthquake takes down the building? That's bad enough, of course, but imagine you attempt to restore from backups and you find that you can't. Or that, because your disaster recovery site was in the same state as your main data center, that hurricane took the DR site out as well. Unlikely as these scenarios may seem, it is both possible and prudent to be prepared.

First of all, make sure you're using the modern platform of choice for your data center: virtualization. Today we have an incredible ability to deploy virtualized workloads literally at a moment's notice. Of course this means that our stakeholders expect the same in the event of a recovery situation. Lucky for us, virtualization has made it affordable to deploy advanced DR and backup capabilities that were previously only available to the very largest enterprises.

Let's take a look at some of the fundamental units of a virtualized workload, things like an e-mail message, a SharePoint item, a file or even an entire VM. Depending on the problem, we may need to restore to any one of those scenarios, and, further, there may be a requirement to restore them off-site. Now, we really have a problem. Or do we?

Talk to any IT professional or service provider who has been down this route before and you will hear a common phrase: "virtualize everything first." Simply put, if the entire footprint of the data center is virtual, then meeting the additional requirements will simply be easier. Everything from protecting workloads off-site to delivering a quick recovery process is made easier when one uses VMs and, this is the second step, data protection tools that are built for VMs. The virtual world is a completely different beast from the physical, so using physical tools won't just leave you unable to take advantage of the unique benefits of virtualization, it will actually degrade your data protection capabilities.

Next, make sure you test all your backups for recoverability. In a physical world, this is far too expensive and cumbersome to do. In a virtual world, it can be done as a matter of course. There's no reason not to take advantage of this capability -- it will certainly save you the heartbreak of an unsuccessful restore after a VM fails.

For some smaller enterprises, setting up a separate DR site or even an offsite backup repository can be too expensive. With virtualization, the costs come way down. Even the cloud becomes an acceptable target for offsite backups. And in a virtual world, it's possible to spin up VMs directly from compressed, deduped backups. So while it's not the same thing as having a mirror failover site, for those organizations that can't afford it, running critical applications from the backup files until you can perform a full restore is definitely an acceptable substitute.

Sure there are challenges along the way such as large data profiles, limited and/or unreliable bandwidth, large virtual machines, critical applications and more. But when those requirements are identified, we can remove the superstition associated with delivering the right data protection approach for these critical workloads.

Friday the 13th doesn't have to be a scary day for IT professionals. If you've prepared properly, even the worst case scenario may not end up being the disaster you used to imagine.

About the Author

Doug Hazelman is the product strategist for Veeam Software. He frequently presents on behalf of Veeam at industry conferences, and shares his expertise via his blog, "VeeamMeUp" and other social media outlets. Doug advises Veeam customers and partners on best practices and key considerations as they implement and better manage their virtual server infrastructures.


Subscribe on YouTube