Does disaster recovery planning matter?

A lot of companies use VDI and RDSH technologies as part of their disaster recovery plans.

A lot of companies use VDI and RDSH technologies as part of their disaster recovery plans, with the idea they can just redirect a bunch of users to some datacenter-hosted desktops if some kind of disaster renders their primary work site unavailable. (Of course I'm not quite sure what kind of disaster would take down an entire corporate campus while leaving the employees state of mind where they actually want to go to work...)

That's actually kind of the point of this article. I was at VDI event a few weeks ago, and during the Q&A after our session, someone asked me what kind of testing they need to do to make sure their disaster recovery plan will work. I thought about it for a minute and finally said, "nothing."

(I'm not saying that I didn't say anything, rather, I'm saying that my answer was the word, "nothing," as in "you shouldn't do anything to test your disaster recovery plan.")

People in the audience laughed because they assumed I was joking, but I wasn't. "Look," I said, "have you ever noticed that every time there's an actual disaster, the disaster plan doesn't work? There's always some excuse about why this disaster was the exception... "Oh, of course we couldn't plan for both of our back haul providers sharing a sewer line," or "Come on... we didn't know the telco's CO was in the basement of the building that burned down," or "How would we know that 25 million people turning on their disaster plans at the exact same time would mean that our cloud provider didn't have capacity?"

Seriously, how many excuses are there? It's always something.

Here's the thing: when disaster strikes, there's always to be something you didn't think about or plan for. This is why it's called a "disaster." If your warehouse is hit by a tornado and your users instantly and seamlessly start working via your backup cloud, then that's not a disaster—that's just called Thursday.

I'm not saying that we shouldn't have alternate plans. (The "Thursday" example is the perfect "disaster avoided due to successful planning.") I'm just saying that we don't need to go nuts with super complex disaster plans that will just fall apart in an actual disaster. (So maybe a consumer NAS sitting under your desk with a continuously replicating copy of all the disk images you have in the cloud is all you need). Just don't get carried away. You can't plan for everything, and time spent trying to plan for every outage is time wasted.

And really, keep in mind that if you have an actual disaster, none of your employees are going to give a hoot about their jobs anyway. They're going to be rushing home to make sure their family is safe or using a garbage can lid to paddle a door down the street to rescue the cat.

(It's like that old joke: "Had this been an actual emergency, I would have run out of the building screaming.")

If whatever knocks you offline is big enough to trend on Twitter, then no ones going to be mad at you for not preventing it.

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

I think there is a slight potential for misinterpretation of your blog that I would like to address -

A lot of IT managers are looking at VDI / RDSH to enable them with DR capabilities in case something happens to the office buildings. IMO it is a good solution.

Paying attention to the article, you were at a VDI event presumably speaking to an audience with running VDI / RDSH environments and for them you say - don't go nuts with over designing.

I'm just concerned that some "headline readers" in a rush might conclude that you do not approve 'Office DR' as being a valid goal with VDI / RDSH.

Or are you? :)

Using VDI / RDSH for office DR is common where I live when unplanned circumstances can hamper or prevent employee ability to reach offices for extended periods.

If (or when) it happens users just work from home or from some company branches or offices at different geographic locations.

For them the best advice would be to work with the virtual desktop as the primary desktop so there would be no surprises in due time.

It is especially important for utilities companies and public sector companies which citizens relay on and you better believe that there are brave workers who will do whatever they can to do their job in times of crisis (Sometime these companies handpick only a handful of business / mission critical employees and give them virtual desktops for that exact purpose).



I find DRP/BCP to be a very loaded topic!

And ... reading this blog post, it made it even clearer for me that there is a huge difference between DRP and BCP!!

I totally agree that in case of an actual Disaster, there are only very very few organizations that need to execute a DR plan (e.g. Police, Fire Dept., Army)

Whereas a BCP is a Good Idea(TM) for everyone in order to keep the productivity up (and the money flowing) when your data center or cloud provider as a little bit of a hiccup (maybe cause by a little flash flood or thunderstorm .... but definitely not part of an actual Disaster)!


Snowmageddon in Atlanta earlier this year is a good example of a "disaster" that shut down a major city and left corporate campuses empty for three days.