Colocation data centers offer the advantages of cost savings, reliability and security through hosted infrastructure for IT equipment. To ensure consistent performance, providers set operational guidelines, or best practices, for tenants to follow when installing their equipment.
However, tenants sometimes view a colocation provider’s best practices for power and cooling as a constraint rather than a useful guideline for IT deployment.
Tenants may feel more comfortable following their traditional deployment design plan, or they may believe that the cost of executing a colocation provider’s best practices will not deliver a measurable ROI. But the manner in which tenants deploy their IT equipment in a colocation environment affects the overall efficiency – and leased-space cost – of the facility.
If a tenant’s IT equipment resides in a “community” that disregards the provider’s power and cooling guidelines, the collective installations may cause stranded capacity – which is capacity that IT loads cannot use due to system design or configuration. This can lead to higher costs for the colocation provider, who in turn, must pass those costs on the tenants.
There’s a simple solution: Follow the provider’s guidelines, which can improve the efficiency of infrastructure resources and reduce costs.
What’s behind the problem?
Colocation tenants and providers both face the challenge of balancing power, cooling and space capacities for servers. Stranded capacity can result from an abundance of low density racks, the design limitations of cooling capacity, or the difficulties of mixing hot and cold air streams. Stated another way, stranded capacity occurs when one type of capacity cannot be used, because one or more other types have been used to maximum capacity.
Approaches to solving the stranded capacity problem include increasing the average rack density – which can address space, power and power distribution issues – and recovering cooling distribution capacity through air containment. But a colocation provider stands a better chance of dealing with this problem and delivering premiere service if tenants abide by the recommended best practices for power and cooling.
Best practices that support rather than hinder
An acceptable use policy (AUP) agreed upon between tenant and provider creates order by spelling out IT deployment guidelines. An AUP controls operational costs for both tenants and providers by setting rules for power, cooling and installations. The benefits add up, improving overall data center performance and delivering cost savings to all parties. Adhering to an AUP can:
- Reduce stranded capacity
- Reduce downtime caused by thermal shutdown and human error
- Extend the life of the initial leased space
An AUP promotes best practices and solutions such as:
- Using overhead cabling
- Using wide racks to accommodate high density cable management
- Preventing mixture of hot and cold air streams
- Using supplemental cooling components for high density racks
Other best practices that benefit tenants and providers include:
- Managing capacity when deploying IT by using data center infrastructure management (DCIM) software.
- Using scalable and efficient powerarchitecture to accommodate rack density increases.
- Using a close-coupled cooling architecture in combination with traditional raised-floor architecture to increase cooling capacity.
Deployment of high efficiency power and cooling technology and management software can help the colocation data center experience be more profitable for both tenants and providers.