NASHVILLE, Tenn. -- Getting money for a data center disaster recovery plan is difficult enough; when the economy goes soft, it gets harder.
That's the word from several attendees of the AFCOM Data Center World show here this week. But by using virtualization, hinting at doomsday scenarios, and calling in favors, some have been able at least to catch the ears of company executives.
Data center disaster recovery plans are often lacking. According to a recent Symantec study of 1,780 data center managers, 33% of disaster recovery plans are undocumented or need work.
Dodging the disaster bullet
One such company is R.G. Barry Corp., a Columbus, Ohio-based footwear company which, at present, doesn't have a data center disaster recovery plan. In a couple weeks, Tom Stoughton, manager of operations and communications, will pitch a disaster recovery plan to his chief financial officer.
"I need to stress … that we have dodged a bullet" thus far, Stoughton said. The company has recently had three downtime events, but they all fell on weekends and allowed Stoughton and his data center staff to recover without major consequences. After two of the three events, Stoughton used the opportunity to tell executives that they needed some kind of disaster backup. Now he's got a chance to make his case.
R.G. Barry runs about 30 servers, all virtualized. Stoughton will use that to demonstrate that his department has "leveraged the IT dollar" intelligently and that DR will be more affordable because of it.
But he's not sure it will work. In a way, he almost wishes the near-misses didn't miss.
"It's almost like I need an event," he said. "I need help."
For those who have implemented data center disaster recovery plans, keeping them up to date can be a logistical and financial nightmare. According to that Symantec study, almost one-third of data center managers haven't re-evaluated their plan over the past year.
Proper disaster recovery requires executive-level support
That often means data centers aren't conducting proper and regular testing. That would be a challenge for a company like Gordon Food Service if Kori Youngblood didn't have executive support. Over the course of seven years, Youngblood, the data center manager at the Grand Rapids, Mich.-based food supplier, built a data center disaster recovery plan at his company.
Fortunately (or not), Youngblood faced a real downtime event -- a three-day outage to the company's storage area network (SAN) -- that pushed the company to put a plan in place.
"You can use these events as a catalyst to drive DR [disaster recovery] work," he said. "Disaster recovery is never easy to get money for."
Gordon Food Service now has a secondary data center about 20 miles away to which it can do synchronous replication. The facility is tested four times a year, which might be a problem with developers if the dictum didn't come from on high.
"Fortunately my CIO and CTO have backed me up every time and said that disaster recovery is a priority," Youngblood said. "That is really what it takes, is getting commitment from the CIO, the CTO, the CFO. It's paramount."
Mark Fontecchio can be reached at mailto:firstname.lastname@example.org.