Towards 'IT as a service'
- Reduce text size Decrease text size
- Increase text size Increase text size
- Print article Print
- Jump to comments Comment
- Share this article Share
- Email article to a friend Email

"Now we’re thinking about the service we provide to the business, not the hardware itself"
The head of Birmingham Airport’s information services department explains how virtualisation has helped its bid to become a service provider to the business
Until recently, Birmingham Airport was one of many organisations where the IT department suffered perception problems. IT staff had a reputation for being highly protective of IT resources and not altogether helpful, a reputation that was sometimes justified.
When Wayne Smith took over the department in 2009, he wanted to change that. He renamed the department ‘Information Services’ and followed the ITIL service management guidelines to become more of an internal service provider than a jealous custodian of precious IT resources.
When much of the airport’s server estate reached the end of its planned lifecycle, Smith oversaw a virtualisation project that brought a number of benefits. It allowed the airport to cut its server estate drastically, and when a water pipe burst in its primary data centre, the flexibility of the virtual system came into its own.
And as Smith explains below, virtualisation chimed with his bid to become a service provider to the business. By removing the cost and effort from provisioning servers, he says, the information services department can be an enabler, not a naysayer.
Not that virtualisation is necessary or sufficient to convert the IT department to a services organisation, but Birmingham Airport’s example demonstrates that the two are at least mutually beneficial.
Information Age: What prompted you to rename your department ‘Information Services’?
Wayne Smith: When I took over what was then called the IT department, a lot of our users saw us as a necessary evil and that our role was to police technology. I wanted to alter that perception, and to change our approach to delivering services to the organisation. We wanted to move away from the phrase ‘technology’ towards services, which implies that we’re here to serve the end-user departments.
Changing the name was a good start. It sent the message to the organisation that things had changed.
How did IT operations change to reflect this services focus?
As much as it was about changing the organisation’s perception of us, it was also about putting over to the staff that internal departments are our customers, and should be treated as such. We trained all our internal staff up on the ITIL Foundation in service management, and most people’s job functions have changed.
We have much more specific service-level agreements now. For example, we didn’t have a dedicated support desk before but now it’s specified in one of our SLAs that we offer 24-hour IT support.
It’s an ongoing programme, and we are now training some key members of staff on some phase two ITIL processes.
Virtualisation
What prompted your recent virtualisation project?
We had a lot of hardware that was coming to the end of its planned lifecycle, including some HP ProLiant DL380 servers. That gave us an opportunity either to replace it for a chunk of money or to look for a smarter way to deliver the same service.
We looked at some alternatives, but virtualisation was such an easy and obvious case to put forward. It’s not in its infancy any more; its tried and trusted.
How did you build the business case?
We approached a number of suppliers besides our existing IT provider; our policy says that we will talk to a minimum of three. We eventually selected Centralis, because their methodology is to scope out the project in advance, with project milestones that are included in the contract. Any cost beyond what is agreed in advance falls on their shoulders, so I know I’m going to get it for that price.
By comparing that project scope to our existing costs, I was able to show that we would save 38% on hardware costs with virtualisation, and that when you included energy and cooling costs the overall saving would be 40%.
What technology is the system based on?
We use VMware for virtualisation and NetApp for storage. The storage was a considerable proportion of the project cost, but we reused our existing servers so that was inevitable. The system is replicated across our two data centres, which are both here in the airport.
How did you go about implementing the virtual infrastructure?
We started with some quick wins, such as line-of-business applications and file servers; things with less dynamic data compared to our transactional systems.
We needed to prove the concept and, because it was new technology for us, we needed some confidence that it was going to work. We left some of the more complex, riskier systems until later in the project, although that was mainly out of caution.
Everything was done in production time, while systems were live. In an airport it’s difficult to find non-production time, because we operate 24 hours a day.
It was all live by the end of November 2010, except for some applications that weren’t suitable for virtualisation such as our flight management system. It is awfully old, and based on a proprietary grid architecture.
What interruption to service was there during the implementation?
It was very quick. Each system was down for a few minutes before it went back up.
The important thing that I wanted to achieve was that I didn’t want the business to know necessarily that a server had been virtualised. We just told them that the server was going down for maintenance purposes. They didn’t need to know [about the virtualisation], and they didn’t notice any difference.
What have the benefits been so far?
We’ve gone from 48 servers down to six. Originally, we’d thought we would have to buy new ones, but we ended up using existing equipment. We had to buy improved network cards, an extra processor here and there, just to beef them up a bit, but that was a lot cheaper than buying six new servers.
Our engineering department has been monitoring our energy consumption, and we’ve able to demonstrate improvements there too.
Next >> Freezing weather puts the virtual platform to the test





