Home > Services > DC > Avoiding Downtime

Avoiding Datacentre Downtime

by John Herbert | Director

Mention downtime, in any context, causes problems for IT managers, typically swiftly followed by sleepless nights.

avoid data centre downtime -  Kelcroft data centre consultants

All too often datacentre facilities fail to perform as expected and one reason is the nature of datacentres. Unlike other facilities that are relatively  static, year in year out, datacentre facilities evolve, it is never a stationary target. A datacentre is just not like other projects, they evolve over time with the equipment regularly refreshed, and a few years after opening it bears little resemblance to the original.

However, the E&M infrastructure is typically static, but needs to flexible to cope with the changing needs, and energy consumption monitored that's why it requires a regular check-up.

It's a bit like like driving your car year in - year out without the MOT (annual roadworthiness test) and then expressing surprise when it breaks down, often at an very inconvenient time. For vehicles there is a mandatory annual inspection, yet the datacentre infrastructure often escapes attention until it is too late, and downtime always hits operations hard.

There are several opinions for the cost of downtime, US$ 1 million hour for a financial institute, perhaps more depending on which way the market is moving. Whatever the exact figure is, one thing is for sure, downtime is not cheap, and the cost of annual auditing is insignificant compared to the energy cost and downtime costs.

avoiding datacentre downtime :: Kelcroft datacentre consultants

New technology including blade servers and the like conspire to defeat the existing E&M infrastructure pressing increasing loads in an environment that was perhaps designed many years before, and unable to cope, the once spare redundancy is used. In case it is needed, the N+1 has been absorbed into the day to day business.

So how can IT managers avoid downtime? an annual health check or audit is essential to manage the risk. After a failure it is too late, the first post mortem question - when was the last check-up?

The industry as a whole seems to agree, a survey conducted late 2006 indicates the primary problems with existing datacentres is infrastructure related. The top five challenges in managing the data centre's listed by respondents:

  1. Limitations on cooling capacity
  2. Limitations of the power distribution system
  3. Achieving sufficient airflow through racks
  4. Assessing future infrastructure needs
  5. Amount of total power delivered from the utility

Managing Risk

In a perfect world we would all have unlimited resources, however the reality is very different. We live in a world of restraints. So how can datacentre operators mitigate the risk?

Kelcroft datacentre solutions - avoid datacentre downtime

The Uptime Institute from the US, creators of the Tier I-IV datacentre classification recommend annual infrastructure audit, in response to demand Kelcroft created the datacentre audit product, the purpose of which is to help you identify and manage the risk.

Owners and operators are involved in the day to day operations, and are often running in 24/7 mode.
An independent review, free from 24/7 pressure, allows a fresh pair of eyes to point out simple fixes to potential downtime.

If you need to manage the risk, and have limited resources a Kelcroft infrastructure audit provides an independent overview of your facility structured to suit your facility.


Further Information

If you need any further information regarding Kelcroft's datacentre consultancy solutions, including data centre, planning, design, peer review, annual auditing, due diligence and thermal analysis contact the experts, call John Herbert at our Hong Kong office, the telephone number is (852) 2335 9830, the fax number is (852) 2335 9862, or simply email us.

Keywords: Avoiding Datacenter downtime