Is Consolidating the Data Center at the Expense of Disaster Recovery?
Data center consolidation appears to be on the increase in 2008, based on newly released research from Computer Economics. Research from Forrester from December 2007 backs this up mightily as a growing trend.
Here are some of the key findings of the CE report :
79% of organizations with less than $350 million in annual revenue are currently engaged in data center consolidation activity, up from 56% last year. The study also found that 68% of medium-sized companies (with $350 million to $1 billion in revenue) and 76% of large organizations (with at least $1 billion in revenue) are researching, implementing, expanding, or have already consolidated data centers. That was about the same level of adoption activity as medium and large organizations reported last year in the company's annual staffing and spending survey of about 200 IT organizations.
My question is, while consolidation is showing an increase, can large enterprises that disperse failover capabilities between data centers afford the risk of single points of failure? Seems like we hear an awful lot from those in the risk management and disaster recovery markets about having escape plans for data and infrastructure, even though the costs of maintaining such environments can be high. (See Baseline's "Disaster Recovery: Two Steps Forward, One Step Back.")
The CE and Forrester research referenced doesn't specifically address disaster recovery, but in the Forrester report, "addressing disaster recovery inefficiencies" ranked fairly high as a key motivation for data center consolidation. Thirty-two percent of 147 of IT executives and managers ranked it in their top five motivations.
How is disaster recovery being accounted for?
I put this question to John Longwell over at CE, who had this to say:
As a part of any data center consolidation, IT managers certainly have to factor in disaster recovery and business continuity planning into the project. Having fewer, larger data centers can increase risks associated with natural disasters. Server virtualization can increase risks associated with hardware failure. But in both cases, you can point to factors where data center or server consolidation can simplify managing disaster recovery as well. Remote backup, redundancy and relocation contingency plans are still requirements. Having two or three data centers may be more optimal than one, but having 25 data centers is not necessarily going to make you any more secure. If you're faced with a situation where a single point of failure is a threat, then you need to invest in dealing with those contingencies. But it's not an obstacle to consolidation. Just a risk that needs to be mitigated by having a plan and periodically testing that plan.
That is reasonable (aside from the 25 data center mention, which seems like hyberole). I get that these things are not mutually exclusive, but I wonder how (especially for the small and midsize players) they are managing the risk exposures.
Frank Scavo, president of Computer Economics, has this to say:
Certainly, any organization needs to take disaster recovery into consideration when considering data center consolidation. No one should consolidate their backup data center out of existence. Yet, as we point out in our study, many organizations maintain multiple data centers not for reasons of business continuity but because they have not rationalized the need for multiple facilities. The principle is, have as few data centers as possible, but no fewer.
Not to be a pessimist or give false cause for alarm, but tropical storm and hurricane season seems to be starting earlier in Florida this year.