Back-up datacentre challenge - to build or to host?

Cliff Saran

Cliff Saran is the managing editor (technology) on Computer Weekly magazine responsible for commissioning, writing and overseeing  the magazine strategy concerning all matters relating to technology from up-and-coming research and development to systems management challenges and legacy support and maintenance.

Cliff has been writing about these subjects since the early 1990s. In his current role, he writes a regular blog called Cliff Saran’s IT FUD blog which aims to unravel the hype, weed out the fear uncertainty and doubt spun by the massive marketing machinery in the IT industry.

You can contact Cliff by emailing [email protected].

View all articles by Cliff Saran >>

How to find Cliff Saran online

Follow Cliff on Twitter

Linked-In profile for Cliff Saran

Cliff's photography website and Flickr photostream.

[email protected] 020 8652 8460 Active Cliff Saran False True

The Royal College of General Practitioners has built its own back-up datacentre using a private cloud architecture, rather than use an external provider.

While the current thinking on datacentres is to using cloud services or hosting for a back-up site, the Royal College of General Practitioners (RCGP) is running its own back-up datacentre, which operates as a "dark" site that goes online when the primary site fails.

After running forecasts of the costs of hosting, Andy Smith, head of IT at the college, determined that the RCGP would be better off running the back-up datacentre itself.

"My five-year forecast showed we would yield a better return on investment if we did it ourselves because we have a rapidly changing environment," he said.

While some organisations choose to run their back-up datacentres in a public cloud, the RCGP needed to maintain ISO 27001 accreditation. "[With a public cloud], I could not guarantee where my data would live," said Smith.

Smith has led an overhaul of IT at the RCGP, encompassing what he describes as a robust disaster recovery plan and simplifying IT support across 22 regional offices. "We needed to look at the scalability for our storage because every time you run out of space you tend to buy more storage."

The college is using HP DL servers running VMware, fibre channel storage using NetApps, and Cisco for networking. Back-up is provided by Dot Hill using Sata drives.

NetApps specialist Vohkus implemented the system, comprising NetApp storage and disaster recovery, with VMware software. This created what Smith and his team characterise as "going from a passive to highly proactive IT structure that's resilient and highly available".

Vohkus installed two NetApp FAS2040 systems located in two datacentres, using Snapmirror to replicate between them.

The new infrastructure has supported the organisation's growth. "Two years ago we were struggling [to support] 36 physical servers. We now have 100 [virtual servers] on 16 physical boxes," Smith said.

Read more:

• Google unveils plans to build flagship energy-efficient datacentre in Dublin
• The impact of the Olympics power drain on London-based datacentres
• Best practice in datacentre transformation
• Internet giants must be more conscious of datacentre greenhouse gas, says report
• Airbus completes high-performance datacentre deployment with HP
• Verizon rejects London datacentre site due to Olympic Games power drain
• Datacentre services supplier Equinix gives its top three tips for apprenticeship selection
• Whitepaper: Building datacentres using containers and modules (requires registration)

MetaKeywords MetaDescription Sensitive Landingpage False