Five Questions to Ask Your Disaster Recovery Service Provider

You’ve done your research, and you’ve looked at different options for disaster recovery (DR). You discarded tapes as slow and unreliable, colo active hosting as too expensive for your budget, appliance-based recovery as too close to your primary systems and therefore prone to the same disaster, and “doing it yourself”… oh, forget about it with your limited IT resources. Ultimately, you determined that cloud is the answer.  While cloud could have been considered a novelty only several years ago, today it is quickly becoming the main stream of disaster recovery. In fact, there are multiple options even within what may seem to be a fairly narrow segment. How do you choose the one that’s right for you? What questions do you ask? What answers should you look for? How do you identify a professional-grade solution? 

Want these answers? Keep on reading…

• How is the solution packaged?

The key here is to understand what other components you will need in order to have the solution that fully meets your needs. For example, does the solution include a full-featured backup function allowing you to restore a file, or a folder, or an email mailbox that you have inadvertently deleted? Some solutions will replicate your servers, but won’t do much to protect your data, at least not in the way a full-function backup solution will. In case of a disaster you’ll need to recover both – your data and your servers. So, look for all-in-one solution.  You will also need to understand what hosting capabilities are available for your replicated environment. Does the provider operate a cloud data center where they can activate a server for you, or will they expect you to arrange a separate target in the cloud (public or private) where your replicated image will live? The best option is to find the provider who can both replicate your server, and store it in an environment where they can quickly (depending on your RTO needs) spin it up for you.

• Do you support my platform?

In a world of heterogeneous environments, it is critical to know the DR provider supports all your critical platforms. Windows or Linux operating systems, physical or virtual environments, different storage options, and even applications – email, databases, directories, etc. – what can you protect easily, without much of new development or integration, out of the box? Even the best provider could be of little value to you if they don’t support your specific configuration.

• What are the data transfer technologies?

How does the data make it from your data center into the service provider’s data center? There are two parts to this question. First, how do you run the initial seed backup? That may involve huge volumes of data – tens or hundreds of terabytes, potentially more. The most effective ways to move such volumes of data over vast distances are not over the wire, but over UPS or FedEx – or whatever the package movers are in your respective country. Does the service provider offer such an option? Most public clouds don’t, which excludes them from our picture. The same considerations apply for failback after the disaster when you need to move the same body of data back into your data center. If your data size is anywhere over 10TB, find a DR service provider that offers a portable (or, rather shippable) data storage device, to transport your data back and forth. Secondly, there is your regular on-going backup and replication. How much data is actually moved every time, and does it fit your backup window? Is it compressed to reduce the network traffic? Is it reduplicated for efficiency, to avoid backing up the same file across multiple volumes? Is it encrypted for security, privacy, and compliance? You may, or may not, need all this. But you need to know the questions to ask.

• Will it work when I need it?

Suppose you are the best planner ever. You’ve thought everything through to the last detail. Every possibility, with no room for error. You had your DR plan reviewed and approved. Except, it all happened 3 years ago. Since then, there have been network changes, hardware and software upgrades, and HR turnover. How do you know your DR will still work? I’ve got three words for you (actually it’s from Forrester Research): test, test, and  test. And how do you test? That’s the 64TB question. Obviously, you don’t want to test on live data (because you may need it), even if it is your backup data. You don’t want to test manually  because it takes a lot of resources for clearly non-revenue-generating activity. Reporting is also important; you want to know what had gone wrong if it didn’t work.  The answer you should be looking for: automated, scheduled, segregated testing. Can you automate your test? Can you pre-schedule it? Does the provider charge for testing? And finally, does the provider give you a sand box, a virtual private testing environment so you don’t play with your live data? 

• Support – will it be there when your IT goes dark?

Support comes in many flavors. First, deployment support. How easy is it to deploy a feature-rich disaster recovery solution? Assuming it involves an on-site component (an appliance) and replication into the cloud, a console for self-administration, designing recovery runbooks and test plans. Even with the latest technology there are many pieces in motion and things to keep in mind. Does the vendor provide a “white glove” support, or do they offer you a “do it yourself” scenario?

Even more important than that, is the support your provider gives you in the event of disaster. When you have a problem, the whole point of the solution and the entire value of your investment is your ability to recover. Are you left alone to deal with it? If the disaster is of a local or regional scale, most of your employees will be dealing with their own mini-disasters (probably the other way around looking from their perspective). Will your disaster recovery vendor stand by your side, holding your hand, making sure you take all of the right steps, or do they do them for you, to mitigate the situation in the best possible way? Will they help you to fail over? Will they support you? And finally, when the worst is over, and you want to safely fail back to your original environment once it is available, will they help you to do that? Keep in mind; even if you have the right DR plan in place, and you have tested it, failing back is a whole different animal. You never know what you are failing back to. So, having a trusted advisor by your side is vital.

Five questions: many possible answers. Pick the ones that work for you. At the end of the day, it’s your internal or external users who will judge whether you had made the right decision or not. But if you know which questions to ask there is higher chance of you getting the professional-grade DR service you need.