If you’re thinking about disaster recovery (DR), you’ve doubtless considered colocating your data center. There are lots of great reasons to colocate, and DR offers one of the best reasons around.
As you consider selecting a colocation provider, it’s easy to get waylaid by some common myths and half-truths that can prevent you from making the best choice. Here are a few to watch out for.
All Backed Up = All Done
There’s a frequent misperception that if you have a solid offsite backup, you’ve basically got disaster recovery nailed. To be sure, colocation provides a great platform for achieving secure remote backups. But if that’s all you have, you’ve missed some crucial steps.
To simplify, disaster recovery can be viewed in three phases.
- First, you lay out your DR preparations. That includes your processes, backup data, applications, systems, network, and other essential ingredients. This is also when you identify the right colocation partner. Getting all of these preparations right will make or break what happens next.
During the preparation phase, you need to define what kinds of disaster are significant to you, based on their potential to disrupt or destroy your business. Weather and geological events, cybercrime, network outages, power and utility troubles, data corruption, disease outbreak – the potential list is long. Some of types of incident will matter more to your organization; others may not be significant risks in your region or sphere of business.
- Second, a disaster occurs. This happens without any effort on your part! If you’ve done your job right, it’s a disaster you planned for.
- Finally — and most important — recovery. Your backup should be ready now, but simply having a backup doesn’t guarantee that you can use it to recover. You need a plan that makes sure the backup survives the disaster and a process to speedily apply the backup to surviving systems with the goal of substantial recovery as quickly as possible.
“Disaster recovery” properly puts the emphasis on its second word. Work with potential providers to ensure that you’ve thought through the recovery process with experts – otherwise your awesome backups will have no place to go.
Farther is Better
You might think that a regional disaster is more recoverable if your colocation facility is as far as possible from your main data center. That’s a myth. Farther may really mean inaccessible – which defeats the whole purpose.
A recent research report from Gartner called “How to Identify the Best Data Center Locations for Disaster Recovery” identifies the delicate balance between far enough to be safe yet near enough to be reachable.
To some extent, the types of disaster you’re trying to mitigate will define the choice of location. If you’re thinking of a regionally focused disaster, such as earthquake or civil unrest, distant locations would appear preferable. According to the same Gartner report, “A secondary site for disaster recovery should be located far enough away from the primary data center to reduce the likelihood of both sites being affected by the same disaster.” There’s no point having the secondary location subsumed in the same regional event that halted your main facility.
However, the tradeoffs of distance are equally clear.
- Farther means slower. When it comes to data transmission, distance works against you. If you’re doing regular, comprehensive backups over a network as you should, latency can cripple your process. The Gartner report observes, “At more than 100 km, even with tuning capabilities, latency issues will begin to impact performance and make any synchronous or active/active recovery model unrealistic.” When it comes to recovery over the network, restoring an entire data center on a slow long-distance line can be monumentally frustrating.
- Farther means less attention. “In addition, if key support staff must travel several hours to conduct routine disaster recovery testing, productivity and inconvenience may negatively impact testing frequency and thereby lessen recovery preparedness,” says this Gartner report. It’s just human nature; your IT team is going to avoid hopping into the car when possible. Smaller fixes may simply be deferred until they get batched with larger ones. And there’s the “out of sight, out of mind” phenomenon that makes the distant facility slightly invisible. Selecting a colocation provider who can manage assignments on your behalf shrinks this disadvantage.
Far is not better than near, nor is it worse. The right facility will be located at the intersection of remoteness and convenience – customized to your organization’s situation.
Just Pick the Right Physical Plant
It surely is essential to find a colocation facility that has the right physical systems: robust security systems, strong power infrastructure and a hardened building. That’s not a myth.
There are excellent, exhaustive checklists of structure, infrastructure and provider attributes that must be considered in selecting a colocation facility. Disregard these at your peril.
The problem comes when that’s all that you look for. In any disaster, it will be the people that dig in and set things right – so your remote facility has to be pleasant and efficient to work in for the human beings you depend on. In a substantial disaster, the remote facility may become living and working space for an entire team for a significant period of time. That requires personal comfort onsite, like good shared and private working environments and human-centered HVAC, as well as local lodging, restaurants, transportation, personal safety, medical support and – if families are separated – communications.
The Gartner report offers a sobering perspective of the serious consequences of a big disaster on team planning: “The recovery team that you think you have, and the recovery team you actually have, may be very different following the occurrence of a major disruptive event given the potential for conflicting priorities and availability.” You can’t overlook human disaster in your preparedness plans.
Selecting a colocation facility for DR is not just about physical attributes. Recovery will take place in a human and societal framework. Overlooking those elements makes the physical systems irrelevant.
Speed is the Only Priority
No myth: the speed of the recovery process is essential to getting your business back on its feet with minimal damage. By all means, track down a facility with top-tier network and system specifications.
But be sure to balance raw speed with the colocation facility’s survivability and operability. During the recovery phase, you may be willing to sacrifice a few micro-specifications for a facility that’s easy to reach from major transportation arteries – which is especially important if your team will be commuting to the remote location. Redundant network suppliers will reduce the possibility of losing your communication lines entirely, making their documented speed irrelevant.
Pick a colocation provider with a stable history and strong track record. That will ensure that your DR partner will be at your side for the long haul, which is a very secure feeling. For optimal longevity, be sure that your provider can convert your colocated DR strategy into the cloud, whether to achieve greater redundancy or simply to allow rapid expansion without pain.
In selecting the right colocation provider for disaster recover, you can avoid falling for some common oversimplifications. DR is complicated; most myths contain elements of truth that you can balance with good sense and forethought.
The outcome? Better DR, of course. And the stability, cost savings, and expansion possibilities that colocation brings to your enterprise.