Five innovative IT projects have been chosen as the winners of the Computer Weekly European User Awards for Data Centre.
The data centre awards aim to recognise innovation in the use of data centre technologies, virtualisation, and its ecosystem of products.
A panel of independent judges viewed entries across five categories: Public Sector Project, Best Technology Innovation, Supplier of the Year, Private Sector Project and Cloud Innovation.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
And the winners are:
- Best Technology Innovation: University of Leeds (Entered by Iceotope)
- Supplier of the Year Winner: Dell (Caterham F1 case study)
- Public Sector Winner: North East of Scotland Shared Data Centre Entered by University of Aberdeen)
- Private Sector Winner: Sudlows
- Cloud Innovation Winner: Totaljobs
A large user of High Performance Computing (HPC) for its research, the University of Leeds was the Iceotope Solution has been installed in a large thermofluid mechanics laboratory at Leeds’ School of Mechanical Engineering, where it is used to perform complex thermal Computational Fluid Dynamics (CFD) simulations. It is also becoming an experimental apparatus for exploring the possibilities of re-using the waste heat from IT systems.
The Iceotope solution of liquid-cooled servers demonstrates a new generation of thinking
Spencer Izard, IDC research manager
While most computer servers use air to cool their electronics, all of the components inside the Iceotope Solution are immersed in 3M Novec Engineered Fluid. The power-hungry fans of traditional computing are replaced by a silent liquid encapsulation cooling process that relies on the natural convection of heat.
At the University of Leeds, the Iceotope solution currently operates contaminant-free in a large thermofluids laboratory, using a total of 107W of pump power to harvest the heat from the IT systems without fans.
Awards judge John Rakowski, analyst and advisor, infrastructure and operations at Forrester Research, said: “The most impressive case study from my point of view here is the University of Leeds using the Iceotope solution. This seems like a great innovation when it comes to green IT and it’s great that the University of Leeds is undertaking this study here.
“Our increasing technology requirements are raising power utilisation levels and, while many organisations talk about green IT, in practice they don't do it because it's costly and means too much change. Projects like this one at the University of Leeds have to be commended as they are investigating real innovative answers to this problem.”
Spencer Izard, IDC research manager and awards judge, said: “As the demands by both academia and business has increased for large compute capability an exponential increase in data center cooling and power management complexity has occurred.
“The Iceotope solution of liquid-cooled servers demonstrates a new generation of thinking and capability in regards to cooling that limits the fallible moving parts of traditional air-cooling fans for a solution utilising cooling fluid for a silent solution utilising natural convection of heat. This in turn reduces power costs as well as providing other innovative advantages not seen in fan based technologies.”
Awards judge Clive Longbottom, founder of Quocirca, said: “The need to deal with heat dissipation in extreme density equipment means that standard forced air systems are struggling.
“Complete liquid immersion is far more effective, and could well be the future as densities continue to increase.”
Caterham F1 Team is a Malaysian-owned Formula One team based in the United Kingdom which travels the world around the year for races and pre-season testing. To support the complex technology needs demanded by the highly competitive F1 industry, the company needed a powerful HPC environment; a resilient trackside IT environment, which can travel the world; and user-friendly tablets with global technology support at the time of need.
The platform is flexible enough to meet the changing demands of a highly mobile F1 season
Clive Longbottom, founder of Quocirca
When the team began its F1 journey, it decided to partner with Dell to support the creation of its end-to-end IT environment and the IT staff at Caterham F1 team. This was no ordinary task though – Dell had to design and deploy an enterprise class environment, including an HPC cluster, in less than 22 weeks.
Following the decision to implement Dell’s HPC solution, Dell consultants held a series of workshops covering the design and implementation of servers, storage, networks and infrastructure applications. This process mapped the team’s short and long-term business goals to its IT strategy. Dell then worked with the in-house IT team to design the optimal HPC environment for Caterham F1 Team. The initial designs were completed on a cloud-based HPC environment provided by Cambridge University, which was later migrated to the Caterham F1 Team’s own HPC platform.
Caterham F1 Team needed the HPC solution to run advanced Computational Fluid Dynamics (CFD) processing. F1 teams have long relied on wind tunnels for testing parts, to find out if they are going to add performance to the car or not. But the relentless advance of technology has seen those teams increasingly use CFD in tandem with wind tunnel work.
Longbottom said this case study demonstrates a true “trusted partner” relationship between the customer and supplier: “Dell works alongside Caterham, not only to ensure that equipment is available, but also to create a platform that is flexible enough to meet the changing demands of a highly mobile F1 season.”
Izard explained that Formula One puts regulatory constraints on the use of compute cycles when designing the cars, as well having very specific in-field/trackside requirements on race day: “Dell demonstrated a strong and pragmatic approach to designing and implementing a data center solution that was both scalable and flexible to the needs of Caterham F1 while minimising the physical footprint and complexity of the IT solutions.
“A combination of compute, network and mobile systems clearly demonstrated an innovative approach to the clients’ needs in locations where IT operations and connectivity is less then optimal.”
The North East of Scotland Shared Data Centre (NESSDC) is a £1.5m initiative involving the creation of a unique shared primary data centre facility for all four tertiary educational establishments in the North East of Scotland.
Whoever was project managing this transformation should be given a bottle of champagne
John Rakowski, analyst at Forrester Research
The project involved extensively upgrading the aging, operational, live University of Aberdeen primary data centre during term-time.
This was a 10-month, large-scale, complex and high-risk strategic project. Utilising, through tender, an external company (Workspace Technology Ltd), and specialist University staff, the major objective in the project was to spearhead a shared-service approach, while delivering significant institutional savings by allowing institutions to vacate current environmentally unfriendly facilities which were unsustainable based on current performance and future requirements. The capital funding for the upgrade project was provided by the four partner institutions.
The project involved re-locating all 400 servers in the University of Aberdeen’s data centre, re-locating and re-wiring 100 live network components and two full Janet racks, replacing all flooring and ceilings and major external construction work.
Unbelievably, unplanned outages across 10 months totalled less than 40 minutes.
Rakowski said this project “is an example of a true data centre transformation”.
“Many organisations would suffer from a number of downtime periods but to only have 40 minutes of downtime in incredible especially with the work that was undertaken. Whoever was project managing this transformation should be given a bottle of champagne.”
Longbottom agreed, saying: “The University of Aberdeen entry shows how a proper approach to shared services can lead to a much more efficient and effective data centre being put in place that meets the needs for multiple users.
Izard said: “Like many organisations in the current economic climate, academic institutions are challenged with providing the best IT services at the best operational price point.
“This entry stood out from the others due to its shared services approach, backed by an energy efficient design, a transition that saw minimal service interruptions, and better use of infrastructure to decrease complexity and operational footprint whilst maintaining quality of service.”
Located on a brown-field site, the BelleVue Data Centre building was set to become an office development but, during the construction phase, was beefed up to enable it to be used as a data centre site.
The physical data centre is the heart of any IT services
This combination, while first appearing too small a footprint, simply required some inventive space planning to fulfil the company’s IT requirement for web-hosting and cloud-based service provision. Once designed, the facility was capable of punching far above its weight in terms of both capacity and resilience, given its modest 400m2 footprint.
The challenge here was clearly going to be the space limitations within the building and the fact that the server rooms only had a footprint of 400m² over five floors. Not only did the building have to host over 160 racks, but the resilience required meant a number of unique technical design features were deployed; all to ensure the facility achieved the hosting company’s own high standards for resilience and energy efficiency.
One of the major challenges came from the fact that the building itself only had a very small, uncommissioned service lift shaft available to the construction team. Subsequently, every individual item of equipment had to be precisely measured and carefully lifted into position on to each floor, without incident.
The roof holds 33 single-circuit air-cooled condensers with fan speed control, plus two condensing units serving both the basement battery room and server room. Sudlows have managed to design and fit 53 Knurr Miracel server cabinets to form highly efficient cold-aisle containment pods on each of the three data centre floors, providing an impressive capacity of 159 racks.
Longbottom commented: “Sudlows shows how clever design and use of the latest technology can create a usable data centre space from a building that at first glance would appear to be ill-fit for purpose.
“Attaining a low PUE is a key requirement as energy prices continue to rise – and free air cooling helps to minimise the energy costs.”
Izard agreed by saying: “The challenge of adapting a building from an office development to data centre capable when the building was not designed for that function is a challenging task.
“This entry demonstrated that, through the likes of in-row cooling and high density data centre design, the challenge of providing a top-end data centre in a restrictive space can now be achieved.”
Rakowski said the description of how Sudlows designed and made its data centre enterprise ready is “inspiring especially with the modern technologies they used”.
“The physical data centre is the heart of any IT services and, with technology based services increasingly fuelling the business, it's essential that their design and build is right.”
Totaljobs Group (TJG) is one of the largest online job boards in the UK. Each day TJG delivers over two million page impressions, 4.5 million emails and processes over 200,000 job applications.
This is a great example of the opportunities of evolving IT operations through the benefits of cloud
Following it purchase by StepStone GmbH in April 2012, TGJ was required to vacate its former owner’s data centres by the end of the year. This was looked on by Totaljobs Group management as an opportunity to build an environment designed to meet ever-changing business requirements while minimising implementation and support costs.
Prior to the sale, former owner Reed Business Information (RBI) had already performed due diligence around the security and resilience of cloud providers and had identified Amazon Web Services (AWS) as a suitable host for their business unit’s IT. Further work by TJG confirmed that AWS could handle their workload without costly re-factoring.
A plan to move the web-facing server estate to AWS while exploiting Office 365 and SharePoint Online for Corporate IT, all in seven months, was approved by StepStone and the fun began.
The work for moving the server estate included upgrading the database environment to SQL Server 2012 and introducing peer-to-peer (P2P) replication, standardising on Windows 2008 R2 operating system and implementing new email, backup and load-balancing solutions.
The new architecture spans two availability zones to improve resilience. Careful design kept additional costs to within 5% over a single implementation, but allowed a full service to be provided if one availability zone fails, and SQL P2P replication allows a failover to be completed in a matter of minutes.
A strong TJG technical team experienced in Agile working practices moved this to a successful go-live in October 2013.
Rakowski believes this case study is a great story of how an enterprise IT organisation can embrace cloud-based technology approaches: “I love the fact that this has not just been a cloud technology implementation but has changed the mindset of the IT organisation.
“Also for such a seismic transformation to have user experience to not be impacted is commendable. Great to see that this cloud adoption does not focus just on cost savings.”
Izard commented: “Combining elements of agile working practices and cloud platforms to evolve a traditional data centre footprint to leverage and evolve into a service-driven cloud system, without a perceivable negative operational impact by the business, while providing noticeable advantages – both from cost and capability perspectives – is a great example of the opportunities, when planned and executed properly, of evolving IT operations through the benefits of cloud.”
All winners will be profiled in full case studies on Computer Weekly. Trophies are on their way to all of the projects mentioned above.