During a presentation he made to an assembled audience of CIOs and data center professionals at Schneider Electric’s Power to the Cloud event, Lex Coors was very clear about the link between standardization of data center processes and infrastructure, and uptime. As Interxion’s VP DTEG (Data Center Technology and Engineering Group) and Chief Engineering Officer, Lex has supervised the design, build and upgrade of over 80,000m² of data center space in 33 locations in 11 countries. Over the past 25 years he has built an exceptionally strong reputation in the design of versatile, cost-effective and energy-efficient data center infrastructure.
I asked Lex about how standardization affects the people operating data centers and how it helps Interxion to maximize uptime. “Think about it – today we have 80,000 nett square meters of IT space in different countries. That means that if you do not standardize – if you had all different types of switchgear from all types of different vendors, all types of different UPSs, with a centralized engineering team to support each and every country you can imagine what could happen in times of crisis.”
“By standardizing on Schneider Electric, for switchgear, UPSs, chillers, CRAC units – even on parts of our BMS systems – we’ve been able to create a situation where we could send people from each country and also from my team, into the Schneider Electric factory for factory training”. So everyone in engineering teams throughout the company gets the same training. Another advantage is that the core engineering team at Interxion headquarters has a direct dialog with Schneider Electric, a conversation which leads to the development of procedures and methods which fit in with Interxion’s infrastrastructure. “Together, we’ve achieved a lot,” says Lex. And that helps to maximise uptime.
According to Lex, human error represents 95 – 97% of all outages of the data center. “If you have a data center which has no component redundancy, that could take down the whole data center. But if you look at the way we design our data centers, it could take out a whole component – which we still don’t like…” he said. “It’s key that you develop training programs with your partner, as we have done with Schneider Electric, that you develop methods of procedures that are standardized. This not only helps us in case there is an issue with a component, but also helps us to understand the Schneider Electric side when they communicate with us… That’s key in a crisis – you need to act quickly and you need to take the right steps. You need to communicate information in a meaningful way.”
So I asked whether there was a clear division – to maximise efficiency you need efficient infrastructure, but to maximise uptime you need well trained people?
Lex said that if you look at data center designs, a lot is built around the failure of a component. “A component failure doesn’t matter as it will not take down a data center,” said Lex. “It might not even create downtime on the SLA. But if a human being does something wrong, say two times or three times in the same row, then they could take out the data center. So it’s crucial that both the partner, Schneider Electric in this case, and Interxion as the co-locator work together and make it into a success from the operations side.”
Schneider Electric white paper 116 “Standardization and Modularity in Data Center Physical Infrastructure” is a free resource which is available for download. Bearing out what Lex has said from his own experience engineering data centers for one of Europe’s leading co-location service providers, it says that failure to adopt modular standardization as a design strategy for data center physical infrastructure causes unnecessary expense, avoidable downtime, and lost business opportunity. Standardization and its close relative, modularity, create wide ranging benefits in DCPI that streamline and simplify every process from initial planning to daily operation, with significant positive effects on all three major components of business value – availability, agility, and total cost of ownership.