Data Center Journal

Volume 35 | December 2014

Issue link: https://cp.revolio.com/i/428453

Contents of this Issue

Navigation

Page 7 of 24

THE DATA CENTER JOURNAL | 5 www.datacenterjournal.com to identify flaws and contingencies it may have missed during the initial planning phase. Woolley cites four areas that com- panies should address in implementing a disaster-recovery plan: • Preparation for and prevention of crises • Determination and classification of crisis severity • Mitigation of the effects of the crisis • Reentry and recovery "e first step to preparedness is un- derstanding the different types of failures that can affect a mission-critical facil- ity: design failures, catastrophic failures, compounding failures and human-error failures," he notes. "Next, data center operators should consider a number of situational assessments including identi- fication and evaluation of hazard prone areas, estimating vulnerability to a disaster and measuring the risk of such a situa- tion. Disaster recovery, fire-safety, security, alarm and backup-power systems should also be evaluated." But preparing for disaster recovery doesn't necessarily mean going solo. Ben- son recommends working with a service provider to deal with the aermath of a cri- sis; doing so, however, also requires plan- ning. "Aer a natural disaster has occurred, there are essentially six steps that facility and data center managers can and should delegate to a service partner: damage as- sessment, inspection and testing, repair or reconditioning, equipment replacement, spare parts support, and acceptance testing and startup." To this end, selecting the right provider is important, as a service call in a disaster is likely to differ from an every- day maintenance request in urgency and difficulty. "You'll want a service provider that has a disaster-recovery plan of its own," Benson said. "For example, it should have contingencies in place if telephone communications are interrupted. You'll want to know how the provider plans to stay engaged with you during an emer- gency, perhaps providing on-site standby support." return on investment Even if you're convinced that prepar- ing for disaster recovery sounds great on paper, you may be wondering about the costs. Naturally, they'll vary in accordance with facility size and uptime needs, market conditions, and what contingencies you're preparing against. For instance, retrofit- ting your facility to be earthquake-proof is a much more expensive proposition than simply planning for utility outages owing to weather events. "Ensuring the data center is secure and an actionable plan in the event of a crisis is in place could save on data center costs and keep the business operational, and thus make it less likely to incur a financial loss due to downtime," said Woolley. "While preparation activities can be initially time consuming, the cost is relatively modest and pays exceptional dividends if a disaster actually occurs." Part of the analysis of potential returns must involve calculation of down- time costs. Assume a disaster strikes and your facility goes down for an hour, a day or a week—or even longer. How much would your company lose in revenue, productivity, future business, reputation and so on? What is the likelihood of such an event occurring, and how much would you need to invest today to stave off these losses? Chances are the price you would pay now to prepare is much less than the price of downtime; and if those numbers are reversed, consider whether you're trying to prepare for too outlandish a pos- sibility (really, why bother planning for an all-out nuclear war?). But not all of the investment need be monetary. If you contract with a service provider for disaster assistance, much of the initial cost will be in time spent researching the options. Although reliance on outside parties should be in addition to your own disaster-recovery plan (some matters only your company can address), it can be a valuable complement. "Having a team of highly trained technicians and en- gineers in place prior to a disaster can help reduce business interruption, get mission- critical equipment back online and restore supporting revenue-generating operations faster than trying to establish this type of relationship aer disaster has struck," notes Benson. other PreParations In planning for disaster recovery, don't forget about the small things. Again, you need not try to prepare for extreme survival situations, but ensuring basic safety and a little comfort for employees in the event of a disaster can make the recovery process much more pleasant. For instance, what if roads are closed or staff is otherwise confined to the building? "Basics such as clean water, first-aid kits, flashlights and batteries are a must have. Data center operators should also keep information such as emergency contact and facility asset lists updated and in an easily accessible location," said Woolley. "In addition, accommodations and provi- sions should be available in the event that the site support staff needs to maintain a continuous presence at the data center. is situation could occur if travel to and from the facility is restricted or hazardous, or if operating with a severely reduced staff complement." Such small (but important) supplies are in addition to fundamentals, like fuel for backup generators. "Because many data centers don't keep spare parts on site, they'll want to ensure their chosen service provider has access to original-equipment- manufacturer (OEM) parts and engineer- ing drawings to streamline the recovery process and ensure adherence to all manufacturer recommendations," Benson said. "Data center and facilities personnel should also consider the geographic loca- tions of the provider's warehoused spare- parts inventory and their average delivery time for emergency parts orders." ConClusions Just keeping up with daily tasks can be difficult enough without the hassle of planning for remote (or, perhaps, not so re- mote) contingencies. Unfortunately, how- ever, the cost of a disaster can be far greater than the relatively minor inconvenience of taking steps now to implement, practice and refine a plan for how to respond should such an event take place. Disasters can and do regularly destroy businesses, but many such failures could have been prevented with a little planning. Ignoring the possibility of a disaster striking your data center or surrounding areas—whether it be a hurricane, earthquake, malicious at- tack or some other event—is a gamble your business can't afford. n

Articles in this issue

Links on this page

Archives of this issue

view archives of Data Center Journal - Volume 35 | December 2014