Co-location Helps Decentralize Data Center Operations

By Lee Kirby  
OTHER PARTS OF THIS ARTICLEPt. 1: Data Center Capacity: Getting the Most from MorePt. 2: Expanding Data Center Allows for Closer ControlPt. 3: This PagePt. 4: Expanding Data Centers to the Cloud Offers More Than Pie in the Sky Benefits

Sometimes the business strategy and operational requirements call for multiple data center sites; however, owning the asset is not desirable from a business standpoint. The benefits of co-location include rapid expandability, which satisfies an immediate need when demand spikes. It is like modularity but without ownership.

The major technological prerequisite to co-location is an application architecture that is capable of operating in a decentralized environment. If the applications cannot run from two or more sites, then they must be re-engineered. That process is more difficult and costly than it may first appear. Larger organizations, in particular, face challenges in accomplishing this because of their legacy applications. These challenges include an application architecture that cannot be run in a virtual environment and distance limitations.

A cost-effective alternative is to compartmentalize, or carve out, a group of applications that do not need to run on multiple sites. For example, move the payroll and human resources applications to a co-location center in Dallas, and maintain the accounting and general ledger applications at the home office's data center in Chicago.

The major disadvantage to co-location is that tenants lose some degree of control over their assets, while retaining the ultimate responsibility for performance. Performance and security are significant risks associated with co-location; therefore, the service provider's ability to meet the owner's requirements are key criteria in selecting a co-location site.

The penalty clause in the contract does not entirely mitigate the security and performance risks. Many co-location service providers offer service credits for outages as a percentage of the monthly lease, not the real and consequential damages to the client's business. For example, a co-location provider might credit 1 percent per minute of outage on a lease of $100,000 per month, or $1,000. However, a two-minute outage can require three to four days of recovery time, the cost of which will far exceed the lease.

If performance is mission critical, then the tenant may wish to identify a service provider that is able and willing to guarantee uptime and cover real and consequential damages for any downtime. This is rare, but it is available, and buyers of these services should use that knowledge to negotiate terms.

Advantage of Triple-Net Leasing

Co-location tenants with larger footprints (typically 3 MW and above) can negotiate a triple-net lease agreement, which make tenants responsible for operating expenses, property taxes and insurance. Because tenants maintain and operate their equipment, a triple-net lease gives them greater control over technology assets, including security and performance.

Optimally, the enterprise customer would be the sole tenant in the triple-net co-location facility. In facilities shared by two or more tenants, there often can be complications as to the demarcation lines — who "owns" what critical assets? — but the installation can be designed to provide separation. If not, either the property management service provider needs to work for all triple-net tenants, or the tenants need to have a clear understanding of their respective responsibilities.

In any case, the enterprise customer must be able to operate separate data center sites and integrate service delivery for maximum performance. This means they must have the staff and operational processes in place to operate separate sites and appropriate management oversight to ensure that all sites are operated to deliver consistent service levels.

— Lee Kirby

Contact FacilitiesNet Editorial Staff »

  posted on 11/16/2011   Article Use Policy

Related Topics: