Dave Aykroid transformed Lake Forest Academy's server and storage infrastructure to deliver high availability.
Aug 13 2009

Earning an 'A' for Availabilty

Lake Forest Academy thwarts downtime with infrastructure overhaul.

Since its establishment in 1857, Illinois' Lake Forest Academy has stayed ahead of the curve in its commitment to prepare students for high levels of achievement in college. In order to maintain its elevated academic goals, LFA must ensure that its technology and infrastructure are fully operational and accessible at all times.

“If our end users are thinking about the availability of servers, we have missed the mark,” emphasizes Dave Aykroid, LFA's director of information technology. “We want them to be thinking about learning and innovating.”

With that in mind, LFA implemented a total availability solution in 2008 that was designed to meet the continuous uptime needs of the school's 100-plus faculty and staff members and nearly 400 students, about half of whom reside on campus.

Keeping Systems Available

“At LFA, total availability means maintaining a server infrastructure our users can rely on,” explains Aykroid. “As a result, our efforts are focused on implementing and refining technology in the classroom, rather than spending our time remediating problems with the infrastructure.”

LFA is now able to do just that, thanks to the deployment of an IBM BladeCenter, VMware Infrastructure Enterprise, Tripp Lite uninterruptible power supply (UPS) systems and two LeftHand Networks SAN storage nodes. The comprehensive solution is earning high marks for facilitating continuous uptime, easing management and maintenance of the academy's critical systems, preserving valuable floor space and increasing efficiency.

“Faculty and students alike expect access to the system at all times, and our volume of activity continues to increase,” notes Matthew Less, dean of faculty and a math instructor at the academy. “This system has handled the increased workload and provided far less downtime.”

Staying Ahead of Demand

Unfortunately, that was not always the case. Describing LFA's previous data-center environment as “a patchwork of aging tower and rack-mount servers,” Aykroid reports that as the school's needs grew, additional tower servers were added.

“Then we found ourselves left with a combination of 14 physical servers of all shapes, sizes and vendors,” he explains. “Heating and cooling requirements grew, and the noise level was high. We were rapidly running out of floor space.”

At the same time, the academy was struggling to accommodate its expanding IT staff. “We required additional desk space,” says Aykroid, “but our data center dominated the largest room in our office.”

LFA was also seeking a more robust UPS system to better protect its critical equipment. “We reside in a historical building dated at more than 100 years old,” Aykroid acknowledges. “Given this environment, utility power can be volatile at times.”

With its previous solution, the academy was also limited to scheduling any equipment maintenance or upgrades during off-hours. “If we needed to do any physical maintenance on the servers, we had to bring them down one at a time,” Aykroid explains.

Storage was another area where LFA's needs were quickly surpassing its capacity. “Our teachers and students are doing great things in the classroom,” Aykroid emphasizes, highlighting the academy's digital music, web design and 3D design classes. “But the large multimedia files created by these classes presented a challenge as our local storage space dwindled.”

As the school began researching different approaches to achieving total availability, IT administrators outlined some key prerequisites for their new system.

“The main thing was having a robust, scalable solution that would be able to keep up with our demands,” says Aykroid. “We want to encourage innovation and productivity among our students, not stifle it because we're concerned about our technology being able to keep up.”

With the installation of the IBM BladeCenter and the move to a virtualized server environment, LFA has significantly improved efficiency and heightened availability. Consolidating from 14 physical servers to just six, the solution has reduced IT costs, improved flexibility and increased energy efficiency.

Management Made Easy

Aykroid has seen a reduction in the cooling costs. “There has been a substantial difference,” he reveals. “The virtualized solution has also made it easier for us to manage.”

For example, the academy can quickly move virtual machines to other servers in order to perform hardware maintenance – without ever bringing down any infrastructure. IT staff members are able to easily patch and reboot servers, as well as rapidly provision new servers and deploy technology.

“VMware's VMotion product allows us to migrate virtual servers to a surviving machine in seconds, permitting us to bring servers down for maintenance, even during the height of the school day,” Aykroid explains. “The level of downtime has been reduced significantly while making our infrastructure more manageable. We can manage the system remotely, and we can restart these servers without having to physically push a button.”

Furthermore, new servers can be provisioned on the fly for staff training or to demo new technology solutions without incurring any hardware costs. “We are able to do more with less,” says Aykroid. “And we can do it quickly. Virtual servers can be created from templates in minutes, allowing the technology staff to deploy these solutions the same day.”

The addition of two Tripp Lite SmartOnline rack-mount UPS systems, which provide redundancy for protection, was another significant step in LFA's attainment of total availability. Each of the 8 kilovolt-ampere units occupies just 8U of valuable rack space, a boon for the academy's data-center space constraints.

To ease maintenance, the UPS systems feature a detachable power distribution unit with a manual bypass switch, enabling hot-swap UPS replacement without powering down connected loads. Aykroid also appreciates the ability to monitor power, temperature and humidity within the data center using a Simple Network Managment Protocol web card.

“The SNMP card sends alerts via e-mail to my smartphone, so I am aware of the environment 24x7,” he says.

To address its storage woes, the academy found relief in LeftHand Networks' NSM 2120, which delivers scalable capacity, performance and availability. The high-density, energy-efficient storage nodes help LFA avoid painful configuration changes such as controller upgrades, expensive software add-ons and paying upfront for future growth, thanks to the ability to scale performance, capacity and availability by simply adding additional nodes.

In addition to the device's flexibility and scalability, Aykroid praises the solution's efficient use of storage. “The SAN allows us to carve out storage volumes the way we want, so we can allocate storage to the area we need it and grow as necessary,” he explains.

Maintaining Power

LFA is among a growing number of organizations that are taking note of the importance of achieving high availability, according to Farah Saeed, senior consultant for the Energy and Power Systems team at Frost & Sullivan. Over the past decade, headline-grabbing events such as the Northeast blackout of 2003, which brought a large section of the North American economy to a standstill, have pushed power quality issues to the forefront.

“These types of events have led to an awareness of what it means to have reliable power, and what a lack of reliable power can do,” Saeed explains. “It can not only affect your day-to-day operations, but can also lead to significant monetary consequences.”

Toward Total Availability

Total availability should include the following:

  • UPS systems
  • Backup hardware
  • Backup software
  • Storage area networks
  • Replication and clustering technologies
  • Virtualization
  • Continuity of operations management software
Credit: Mark Battrell