When the firm’s IT department set out to modernize and consolidate their server infrastructure -- which consisted of 60 rack mount servers equipped with direct attached storage -- they had a number of objectives in mind.

The first objective was to provide a faster, more responsive and more robust user experience. Andy Beech, the firm’s IT Operations Manager, commented, “Our previous infrastructure was hardwired. Each server had direct attached storage, ran a specific application and was cabled into specific networks. Consequently, it was very difficult to match resources to workloads. We knew that server virtualization would help, but we also wanted to ensure that our infrastructure would not limit either performance or flexibility.”

A second goal was to save space. As the firm grew, every square meter within their central London data center had been consumed. Because this was some of the world’s most costly real estate, it was critical to accommodate the firms’ rapidly increasing demand for IT services within the existing footprint.

The third objective was to establish a disaster recovery capability. “We knew we needed a DR facility for our main data center,” Andy continued, “and we wanted a flexible infrastructure that we could manage remotely to as great an extent as possible.”

In planning the project, it was clear that server virtualization would be one of the project’s pillars. But it was also clear that a traditional infrastructure would be inadequate to meet the firm’s continuously changing needs. “We get a lot of short notice requirements,” said Peter East, Farrer’s Infrastructure Supervisor. “And client requests now have a much higher information technology component than in the past. More collaborative services and easier access to attorneys and information are now the rule. We need to respond quickly when these new demands arise.”

To meet these agility needs, the IT team considered several different options, including virtual I/O technology from Xsigo. “Only Xsigo provided the combination of attributes we wanted,” continued Peter. “When we evaluated Xsigo’s fabric, we found that it was fast, resilient, and easily scaled. In addition, we found the highly intuitive Xsigo management environment met our objectives to simplify management and increase agility.”

“We also considered a traditional 10G Ethernet infrastructure. This might have been an obvious choice since we had decided to implement a Compellent iSCSI SAN, but we found that Xsigo delivered more performance for the money, enabled far more management agility, and made it easy to integrate the SAN. Furthermore, with Xsigo we could configure physically isolated network and storage connections without relying on VLANs, a valuable assurance for us when handling confidential client data.”

In the end, two Xsigo I/O Directors were deployed at Farrer’s London data center, along with HP DL-series servers, VMware virtualization software, and a Compellent Storage Center SAN. A scaled-down version of the same infrastructure was also deployed at the firm’s DR facility.

The result was a more easily managed, more agile data center. Recalling an occasion where agility provided immediate benefits, Andy Beech commented, “Our practice management software needed fault tolerance capability, a requirement we wanted to meet with VMware’s FT feature. But this required a separate network for the VMware FT traffic. With a traditional infrastructure, this would have required several weeks to deploy and configure. With Xsigo, we had a fully isolated network
up and running in five minutes.”

The Xsigo environment also enabled a more efficient workflow for implementing changes such as software updates or new service rollouts. “We are now able to develop streamlined processes for testing and staging by creating segregated areas of the network where new services or upgrades can be developed,” Andy concluded. “With traditional infrastructure, this was very difficult to achieve.”

Xsigo’s high I/O bandwidth further boosted efficiency and space savings by delivering the needed compute capaciy with about 1/3 fewer servers than would have been required with conventional I/O.

In the future, the IT team plans to increase the capability of the disaster recovery site. This will allow the compute capacity at the DR site to augment the main site, adding needed capacity during day-to-day operations. Xsigo virtual I/O will allow rapid infrastructure reconfiguration if needed for DR purposes.

Andy summed up the project stating, “Our Xsigo implementation went very smoothly. The Xsigo team was extremely helpful, and the training was quite clear. At every stage, the support we have received has been top drawer. The Xsigo engineers understand our issues and respond promptly. It’s refreshing to deal with a very professional support team that is always there and always pulls through. The experience has been very positive.”