How to choose the right data centre
How to choose the right data centre is explained by KRIS KUMAR of Digital Realty.
For large corporate enterprises and organisations operating in industries such as financial services, IT services, cloud/managed services and/or telecommunications, the cost of a disruption to system availability in data centres can be enormous. Not only in monetary terms, but also in terms of its impact on market share and corporate imperatives.
Companies need assurances that the strategic data centre solution they choose can live up to their stringent requirements around uptime, reliability and security. And, third-party certification, such as Uptime Institute’s Tier Certification, provides confidence that a data centre being considered for critical business operations has been reviewed and validated.
ENSURING SECURITY, RESILIENCE AND EFFICIENCY
The Tier certification process is performed in two stages. First, engineers evaluate the design topography and equipment facilities of a data centre to ensure the underlying solution is Tier compliant. This is followed by an on-site inspection and validation process to ensure that the facility has been built and commissioned according to the submitted design documents.
The design topology may be optimised depending on a country’s climate. In Asia Pacific, for example, while data centres in Sydney and Melbourne may be designed with roof-mounted, free-air cooling systems suited to a dry climate, data centres in Hong Kong and Singapore may use chilled-water systems better suited to the tropical heat.
Rigorous uptime requirements and long-term viability often drive organisations to select the strategic solutions found in a Tier-certified infrastructure. As a minimum requirement, many companies need their site infrastructure to be ‘concurrently maintainable’ as defined by Tier III, which allows for planned maintenance or emergency repairs of systems to be conducted without disrupting operations.
Increasing levels of component redundancies (N+1 and 2N configurations) enhance a data centre’s availability and consequently the reliability of the facility. A Tier III certified data centre has redundant capacity components (N+1) and multiple independent power and cooling distribution paths designed to exceed 99.982 percent uptime. As only one path is active at a time, routine maintenance and some unplanned outages are not expected to have a negative impact on data processing equipment.
While many providers claim to build Tier III or Tier IV data centres, the only way for customers to verify this is if providers actually undergo the rigorous certification process. As such, tier certification has become a valuable tool to help customers evaluate and differentiate between potential partners, and it is increasingly becoming a customer requirement in the procurement process.
Investment in the certification process is indicative of a data centre provider’s commitment to delivering the high standards expected by customers and it also affords an opportunity for a provider to demonstrate its design credentials.
THE ‘HOLY GRAIL’ OF SYSTEM AVAILABILITY
Some providers have managed to achieve what is considered the ‘Holy Grail’ of system availability — the five nines, in other words, 99.999 percent uptime. In relation to industry claims of Tier III, this or ‘five nines’ that, a customer charged with making crucial decisions about which provider may potentially house its mission critical technology infrastructure, it needs to see evidence.
If a provider claims to build to a Tier III or Tier IV standard, ask to see their certification. If they claim a four nines or even five nines uptime, again, ask to see evidence of this. Any reputable data centre provider should only be too happy to share details of their certification and uptime data credentials.
Kris Kumar is senior vice president and regional head for Asia Pacific at Digital Realty.