I-Business Network, a financial/accounting software-as-a-service (SaaS) provider located in Marietta, Ga., faced both of these challenges head-on. I-Business Network had implemented a hypervisor used primarily for test and development. The production side of the company's data center dedicated individual application servers to each client.
The low utilization of each server, coupled with the growing maintenance and management requirements imposed by a physical infrastructure, made server virtualization an inevitable choice for consolidation and management. But when Ty Hacker assumed his role of director of technical services at I-Business Network, he had serious concerns about efficiency and utility of the company's current hypervisor technology. Limited deployment also meant limited staff expertise.
Requires Free Membership to View
When you register, my team of editors will also send you the latest expert resources covering all areas of server virtualization, such as platforms, architectures and strategies, server hardware, managing virtual environments, application issues and more.
Cathleen A. Gagne, Senior Editorial Director
 |
||||
|
|
|||
|
||||
To achieve desirable levels of server consolidation while supporting application streaming to clients -- a mix of both server and application virtualization -- Hacker leveraged his experience with an alternative virtualization platform.
"I had that experience under my belt already, so I immediately knew what had to happen in terms of moving forward with server consolidation," he said. The initiative also drove the implementation of SAN equipment and other infrastructure improvements, Hacker noted.
Lab testing played an important part in the transition. Hacker spent about three months preparing for the newly planned server and application virtualization platforms, paying particular attention to the associated API used for optimization. Testing revealed the new application server platforms would yield about 97% of the performance experienced in traditional bare-metal systems.
"For a company hosting business applications, it had to appear as though we made no switch at all -- not only on the hypervisor, but on the SAN as well," he said.
Hacker also reported that the transition to the new hypervisor required no new server purchases or upgrades. This was largely because workloads were not CPU-intensive. "We repurposed 60% of the servers that were in our data center," he said, noting that the deployment involved new presentation servers, new virtual machines, SAN implementation and server consolidation. Repurposing a majority of servers saved the company tremendous capital.
In the future, Hacker will follow the new virtualization product roadmap, but he also intends to investigate ways to incorporate other hypervisors into the environment where they might make the most sense for specific workloads.
This was first published in December 2009