Ever since we first started designing it, there’s been a great deal of focus on maximizing efficiency at the new Brocade San Jose campus. From the outset, we’ve put considerable thought and effort into how to contain space, power, cooling, and management costs. In terms of operational costs, servers are both a significant portion of the IT spend for Brocade and a good way to drive efficiency.
Brocade IT has standardized on blade platforms for all of our X86 processing needs. Although replacing individual legacy X86 platforms required investment, the efficiencies gained through standardized management—lower cooling costs, better power and space management—are expected to deliver significant OpEx reduction and long-term cost savings. Another underestimated bonus of blade platforms is physical footprint standardization. On the face of it, having the “same size box” doesn’t seem to be a huge advantage, but once you consider issues like rack and cable plant planning, hot-aisle containment, and installation simplification, standardization has a significantly positive cost impact.
With the new campus deployment, the “heavy lift” server for monolithic workloads is the Sun M Series platform. This platform’s virtualization capabilities, coupled with SAN storage accessed via Brocade HBAs, provides a very stable foundation for code management systems and business applications. Migrating from a legacy environment with multiple UNIX variants into a single form factor with virtualization should improve the availability and performance experienced by engineering staff. Development code is replicated to and from Brocade HQ to multiple Brocade development sites worldwide, so performance and availability of these systems is paramount.
Virtualization has become the standard within Brocade, and a large part of the campus migration has been from physical to virtualized servers: 63% of our servers in the campus are now virtualized. The benefits of virtualization are well known and we’ve applied them to both our legacy and current needs. For instance, our customer and OEM partner relationships require long-term support of many legacy products. To back this support commitment, virtualization has allowed us to support multiple generations of legacy development environments without the corresponding plethora of physical servers that would otherwise have been needed. In addition, we can meet our current application needs with a reduced number of physical servers that each provide a platform for multiple VMs. Our virtualization approach is not dogmatic, and we use the best solution for the need at hand – e.g., Microsoft Hyper-V for Windows, and Sun Solaris Containers for Solaris systems and VMware for Linux systems. Moreover, virtualization is enabling Brocade IT to further its “cloud” computing strategy as well as lay the foundation for more effective business continuity solutions.
It should be stressed that the key component to making all of these standardized platforms a reality is centralized storage. One common factor in choosing blades and the Sun M series was that the systems, by design, only have sufficient storage to boot an operating system (if at all). This leaves all application and business data on centralized SAN and NAS storage and avoids the “islands” that occur with DAS. What makes all this storage viable is a robust high-performance networking infrastructure that utilizes the full range of Brocade SAN and IP products (data center, enterprise, and carrier).
Employee move-ins started last month and about 50 percent of all San Jose staff is now onsite at the new campus. And, with very few exceptions, it’s been a smooth transition. From a business continuity standpoint, the rigorous attention to detail throughout the planning and design phases of the new campus is now paying dividends. Thank you for your continued interest in the new campus and Brocade IT’s contributions.