From a data center perspective, the last five years have been very interesting. We've seen a huge push by customers of all sizes to reduce costs and improve data center performance and resiliency.
In the process, we get called on to assess whether existing network architectures can meet the requirements, or whether a transition to next-generation architectures -- for example, Cisco's Unified Data Center -- would be best.
[ BY THE NUMBERS: Driven by 10G, Ethernet switches buck economic headwinds ]
The first order of work is to understand the customer's business and technical needs. For example, many companies have opted to move from physical servers to virtual servers. Over the last four to five years, server virtualization has been driving transitions from 1G to 10G Ethernet. With that, we have seen potential issues with oversubscription at several places in the network -- particularly in the data center core, where there may be a lot of servers, and in aggregation, where multiple locations come together. We have also seen challenges associated with existing architectures and workload mobility.
Consolidation is another significant factor to consider during virtualization. Due to 10G growth, relying on an organization's existing architecture during consolidation usually requires not only a supervisor and linecard uplift, but in some cases, more chassis -- and those additional chassis come with higher costs for space, power, cooling, cabling and maintenance.
Next-generation example: Cisco's Nexus 7000 platform
Requirements will vary from one organization to the next, but the beauty of a next-generation solution such as Cisco's Unified Data Center, in particular incorporating the Nexus 7000 platform, is that it's designed with Layer 2 switching/Layer 3 routing modular performance, making it especially great for virtualization, consolidation and redundancy, even in a dense, high-performance data center core.
[ TECH ARGUMENT: Cisco Catalyst 6500 vs. Cisco Nexus 7000 ]
Essentially, it enables your trusted solutions provider to design a cost-effective, highly available, resilient, 10G wire rate collapsed architecture. It is important for organizations to understand the value of a solution such as the Nexus 7000 platform, but perhaps even more important to understand that this type of solution can provide organizations of any size with a smoother transition during data center consolidation and virtualization.
Operational efficiencies are just as important as performance in the data center. Customers that opt for a next generation solution will have operational flexibility in the same way that they have a choice of Internet operating systems. For example, Cisco's Nexus 7000 platform offers:
- Highest availability with Hitless ISSU (In Service Software Upgrades), Hitless Supervisor Switchover, Hitless Fabric Online Insertion and Removal (OIR), Stateful Process Restart, Graceful Restart (GR) and Non Stop Forwarding (NSF)
- Protocol Fast Convergence in hardware for Border Gateway Protocol (BGP), Intermediate System To Intermediate System (IS-IS), Open Shortest Path First (OSPF), Enhanced Interior Gateway Routing Protocol (EIGRP), TE-RFF
- Gold, EEM, embedded Wireshark, Flexible Netflow; open XML Interface, integration with Prime LMS and python scripting
- New NX-OS capabilities, such as Port Profile, CONFIG-SYNC, and module pre-provisioning enhances management
- Integration and validated designs with Adaptive Security Appliance (ASA), Application Acceleration with WCCP (Web Cache Communication Protocol)
- Service chassis integration with LACP (Link Aggregation Control Protocol) between N7000 vPC (Virtual Port Channel) and Cat 6500 VSS (Virtual Switching System) making migration to Nexus 7000 aggregation seamless
Next-generation solutions also support innovation with their advanced features. Moving to a solution like Nexus 7000 can remedy real-world challenges attributed in large part to virtualization and consolidation in rapidly growing and evolving data centers, with features like:
- Virtual Device Context (VDC), which enables organizations to virtualize their switch for Collapsed Core/Aggregation, R&D, DMZ deployments. The Nexus 7000, for example, supports true chassis virtualization. Essentially, you can allocate memory to each virtual device context, in addition to CPU shares and configuration files -- all while maintaining flexible separation/distribution of hardware resources and software components, data plane and control plane separation, software fault isolation and, very importantly, securely delineated administrative contexts.
- Overlay Transport Virtualization (OTV) with Locator/ID Separation Protocol (LISP) are other advanced features that can solve workload mobility issues. Though easy to deploy, planning is important, since OTV-LISP enables customers to extend mission-critical applications across any transport between data centers. The planning is key when considering which applications are mission critical and on which VLANs to extend. Once OTV and LISP are running, organizations can benefit from dynamic real time workload mobility and disaster recovery across multiple data centers.
The list of advanced features for next-generation solutions such as Cisco's Unified Data Center and the Nexus 7000 is longer than we can discuss in this article, but again, it is all about your organization's unique needs. Today's data center solutions are dramatically different from what was available five years ago, and they are designed to address specific data center challenges faced by organizations of all sizes. Take some time to consider what they might do for yours.
About CDW: CDW is a leading provider of technology solutions for business, government, education and healthcare. Ranked No. 31 on Forbes' list of America's Largest Private Companies and No. 270 on the FORTUNE 500, CDW was founded in 1984 and employs more than 6,900 coworkers. For the trailing 12 months ended September 30, 2012, the company generated net sales of more than $10 billion.
Read more about pc in Network World's PC section.