Issue link: https://cp.revolio.com/i/1012043
16 | THE DATA CENTER JOURNAL www.datacenterjournal.com eter. In addition, voltage readings on UPS and PDUs should be recorded and available in an easily accessible repository. FUEL, FUEL AND MORE FUEL Natural disasters oen interrupt power to buildings or regions. A solid data center will keep enough fuel on hand to run operations for at least 24 hours. Before a storm hits, for example, data center operators should top off all fuel and run generators with load banks to ensure they're in good condi- tion and ready to go in less-than-ideal conditions. All generators must be in automatic mode so that when power goes out, they automatically kick in and run operations without interruption. In addition, data center opera- tors should have multiple fuel-vendor contracts and should alert fuel ven- dors ahead of storms. If a blackout or unexpected disaster strikes, they should maintain at least 80 percent fuel capac- ity and consider installing supplemen- tal tanks to ensure fuel is flowing to the generators to keep mission-critical business applications running. It's also important to understand on-site fuel restrictions for each municipality. In addition, data center operators should require fuel vendors to include them on their emergency routes so they're among the first organizations to receive fuel. TRAIN IN ADVANCE Finally, training is a vital part of overall disaster preparedness. Good data center operators should simu- late failure scenarios from equipment breakdown to power outages to security breaches. Additionally, they should run and document disaster-preparedness drills at least once a quarter. When pos- sible, colocation and public data centers should invite customers to partake in these drills at least twice a year to better understanding how the facilities will manage a crisis. REVIEW THE PLAN Organizations should regularly re- evaluate their data center's disaster-re- covery plan and know the contingency plans, including redundant operations, in case a natural disaster strikes. With proper planning, they can minimize downtime, effectively navigate both foreseen and unforeseen outages, mini- mize financial impact, and protect their reputation when disaster strikes. n About the Author: Ross Warrington is a regional vice president of operations at vXchnge and is responsible for managing all 14 data center locations. With more than 30 years of experience, Ross has managed data center construction, engineering, repair and maintenance, leading him to the emerging business of colocation. He has participated in colocation design, construction, engineering and operations in his various roles. He is a Vietnam Era veteran, having served in the U.S. Marine Corps from 1969 to 1973. SAVE THE DATE 2018 FALL CONFERENCE END-TO-END RELIABILITY: MISSION CRITICAL FACILITIES 2018 FALL CONFERENCE END-TO-END RELIABILITY: MISSION CRITICAL FACILITIES OCTOBER 21-24, 2018 JW MARRIOTT DESERT RIDGE, PHOENIX, AZ OCTOBER 21-24, 2018 JW MARRIOTT DESERT RIDGE, PHOENIX, AZ METRICS OF THE FUTURE METRICS OF THE FUTURE www.7x24exchange.org RICH KARLGAARD Publisher Forbes TECHNOLOGY AND 2018 TURBULENCE CONFERENCE KEYNOTE www.7x24exchange.org/linkedin @7x24exchange #7x24exchange 18Fall MC_STD_8.375x5.25_3fxgrx.qxp_DM 6/18/18 2:16 PM Page 1