This week, I participated in the Wikibon Peer Incite entitled "Creating a Zero Data Loss Environment" during which Tim Hays, VP of Information Technology at Animal Health International, discussed his company’s implementation of Axxana – essentially a data center "black box" technology – in order to create a zero-data-loss environment. The discussion left clear takeaways that make such endeavors interesting and, in the right circumstances, almost a no brainer.
Axxana has created a product that leverages EMC’s RecoverPoint technology to create what amounts to a zero-data-loss situation. With the Axxana and RecoverPoint combination, zero data loss can be achieved even over long distances.
While the Axxana solution currently requires RecoverPoint in order to work, let’s consider some of the opportunities that such a scenario might provide. First of all, when it comes to disaster recovery, the sheer cost of implementing the perfect scenario is often prohibitive. When one considers disaster recovery as nothing more than an insurance policy, organizations have to make tough choices about how much insurance they’d like to buy. When we spoke this week with Mr. Hays, as he compared a number of different complete disaster recovery options, he described the Axxana solution as a “no brainer” when compared with the cost of what it would take to implement a three-data-center solution. In fact, the massive communications costs alone would have been far more expensive than what Animal Health International ended up paying for the Axxana solution.
When CIOs read papers regarding these kinds of initiatives, the general guidance is to perform a number of analyses, including a business impact analysis, which leads to decisions around appropriate recovery point objectives and recovery time objectives. Mr. Hays was extremely fortunate. The zero data loss solution is about as good as it gets. So, if the cost differential is so insane that it becomes a no brainer to go with the best insurance possible, do it. This was the situation that Mr. Hays found himself in.
Personally, I live those kinds of decisions! While it’s great to follow the “best practices” and expend a ton of effort, when you don’t have to bother, it’s wonderful.
Further, when it comes time to determine how much insurance you want in your disaster recovery plan, the increasing complexity of modern data systems can make this analysis incredibly difficult. More and more, it’s incredibly difficult to completely decouple systems from one another. Applications, processes and services are so intertwined that attempting to make granular decisions about which applications need to be considered mission critical can be difficult, if not impossible. Further, this application tangling adds much more complexity on the recovery side of the equation, particularly if you attempt to decouple them to keep DR costs down.
What if you could just forget having to think about what should and should not be considered mission critical? What if you could just say “protect it all” and be done with it? What if you didn’t have to worry about tangled apps? Because of the cost-effectiveness of the Axxana solution, Mr. Hays was able to do exactly this. He’s protecting both his production and development environments with the solution and is doing so in a complete way.
There remain a great number of organizations out there that have either poor or no disaster recovery plans. The reasons vary, but two big factors are cost and complexity, the two issues that Mr. Hays has been able to eliminate as factors from his decision.
There is a common theme here that must be observed: Simplicity. Mr. Hays was able to eschew what is often considered the “best practice” with regard to disaster recovery and, rather than have to spend weeks developing complex business justifications metrics and making decisions around what could and could not be “insured”, use a simply back-of-the-napkin cost/benefit approach and say "protect it all." Again, this is not due to laziness or lack of judgment... in fact this is due to the no-brainer aspect of the zero-data-loss solution he chose.
Of course, not everyone will have this luxury. If you’re not an EMC customer or don’t want to be locked into EMC for whatever reason, today’s Axxana solution is not for you. That said, there are other solutions out there that may prove to be just as obvious as the one chosen by Mr. Hays. By moving forward with this simple, elegant solution, Mr. Hays’ team is able to spend much more time on the truly important side of the disaster recovery equation, which is recovery. Further, the time that might have been spent planning and constantly testing a traditional complicated disaster recovery solution can be spent on business-facing value-add initiatives instead.
Action Item: That was a major take away for me this week. Simplicity. It’s so easy to overengineer and overthink solutions and create structures that are barely sustainable. When possible, go for the simple; go for the elegant. Go for solutions that make you say, "Well, duh!" and spend more time making sure the business remain operational and gets better.