Are You Prepared for Disaster (Recovery)?

Most companies cannot efficiently execute their disaster recovery plans. Surprised? Nervous? If you can see your company in these murky waters, you should be…

Tony Beeston
Tony Beeston, June 30, 2017 10:00 am
Blog > Disaster Recovery | ASO | Service orchestration > Are You Prepared for Disaster (Recovery)?

There’s no escaping how essential IT has become to modern business; gone are the days where corporate life can continue without its IT systems. These days, across all industry sectors, critical business processes rely upon IT, and yet we’re still being met by what feels like an age-old conundrum: what awaits us in the face of a disaster?

Just to be clear, when I say ‘disaster’, I’m not talking about a situation which can be solved by the likes of high availability. In this case, disaster means a ‘marginally-short-of-an-apocalypse’ scale incident. Simply load-balancing servers isn’t going to solve this situation. Before you get too comfortable and tell yourself, “that sort of thing will never happen to me”, I want to run some figures past you quickly, which have been compiled by IDG Group:

  • 42% of surveyed companies experienced a ‘catastrophe-marginally-short-of-an-apocalypse’ event in the last year alone
  • 65% of firms are still relying on manual, human factors in their disaster recovery plans
  • 72% of companies only test their disaster recovery plan (DRP) once a year

So, how confident are you now that you won’t experience the same sort of issue?

There’s no accounting for the ‘when’ or ‘why’ once disaster strikes. It could be caused by human error, industrial sabotage, technological failure, or even an act of god – the root of the problem at this stage is irrelevant. As you know, businesses rely on IT implicitly, from communications to point of sale. In the immediate aftermath, the cause of the disaster is a secondary concern; your number one objective is simply how quickly you can get your systems back online and return to normal operations. What you need is solid, fool-proof DRP execution.

The Problems with Disaster Recovery

Historically, planning for disaster recovery (DR) has been problematic, to put it lightly. It’s a time-consuming and costly process. A good DRP will work in tandem with an organization’s business continuity plan. Critical processes, procedures and programs should be highlighted as such and each should be attributed a recovery time objective (RTO) and recovery point objects should be outlined. In conjunction with the business continuity plan, the DRP should indicate a timeframe for an expected return to operation, and herein lies one of our problems.

As our systems have become more complex, with ever increasing amounts of inter-dependencies, DR documentation has become unwieldly. So many systems are interconnected and dependent upon each other in manners previously unthought-of, deciding which systems are and are not critical is cumbersome. A good DRP should be brief and easy to understand. An epic, 400-page document is of no use to anyone, especially in the face of a real disaster – on the other hand, neither is a hastily scrawled note on the back of an envelope.

Moreover, DRPs should cover all eventualities, from major catastrophes like earthquakes and flooding, to smaller issues such as power failures.

Just as we have more and more interconnected systems, the state of IT today focuses on the ability to adapt and evolve at the drop of a hat. Such change must also be accounted for within your DR site and plan. However, just trying to keep up with all the changes, let alone document them is a logistical nightmare. How can you be sure all contingencies are accounted for?

However, having a DRP is but one step in the process. It’s all well and good having an immaculate document but without ever putting it to the test, it’s still only a theory. Periodically testing will enable you to understand how it will be executed. It will highlight previously undetected deficiencies you couldn’t afford to discover in a real-case scenario.

Yet companies aren’t testing their DRPs frequently enough. They’re testing only once or twice a year, finding flaws and never really taking any proactive steps to rectify these. Many companies are still almost entirely reliant upon human factors and manual input, which introduce the very real possibility of errors occurring during a real DR situation.

FEMA has estimated that up to 40% of businesses won’t return to operations after a disaster. Deficiencies in DRPs and testing methods play a big part in making up this number. Devising an efficient DR strategy is complicated, without a doubt. Luckily, when it comes to putting the plan into action, there’s a more streamlined process. 

The State of Disaster Recovery Today

Given enough time, everything evolves, including DR. While we used to rely upon manual implementation to facilitate and action disaster recovery plans, that doesn’t have to be the case anymore. Like any repeatable, systematic action, we’ve devised automation processes to remove the element of human error.

Automating your DRP ensures that every time you execute (real or test), the same actions will run in the same order and the same checks will be initiated. Automating your DRP removes variables from the equation. You’re no longer reliant upon human factors, factors which could fail under the pressure of a real implementation. An automated system ensures the people involved in the process cannot make an error. Your DR is now entirely auditable and consistent, delivering proof of capability which can be easily documented. System upgrades and additions can quickly be added to your workflow, meaning you don’t have to worry about overlooking potentially business-critical system dependencies.

The biggest advantage of an automated approach is it enables us to test near-enough at will. Being able to drill so easily and quickly provides the peace of mind that if a ‘catastrophe-marginally-short-of-an-apocalypse’ event does impact us, we’re not going to be met with any unaccounted-for surprises in our fail-over procedures.

Of course, some manual sanity checks will still need to be implemented, but with an automated system human involvement has been minimized as much as possible. Without the reliance upon manual procedures, automation and modern virtualization techniques mean businesses can be back and operational within minutes, rather than hours or days. It gives us the confidence that a disaster will not compromise our business. Does your DRP promise that?

Enterprise Disaster Recovery Automation & Monitoring

Suggested resource

Enterprise Disaster Recovery Automation & Monitoring
Datasheet Download

Enterprise Disaster Recovery Automation & Monitoring

Discover how Automic Disaster Recovery eliminates the risk and complexity of DR execution, and can help reduce the costs of operating your DR environment.

Disaster Recovery
Service orchestration
Back to the blog
Tony Beeston

Tony Beeston

Tony Beeston is Product Marketing Manager at Automic Software, and has over 30 years’ industry experience. For the last 15 of these he has been delivering solutions for running end-to-end business processes across ERP systems to companies worldwide.