Maintaining business continuity as data centres transform

An Industry Insight feature with HP ProCurve

For enterprises today, the data center is the epicenter of business. But what do you do when this crucial nucleus is in a state of flux? HP ProCurve’s CTO Paul Condgon reveals how to keep your data centre on track at all times.

“Visibility means collecting and filtering the right kind of data to allow for trends analysis to find high-risk areas”
-Paul Congdon, HP ProCurve CTO

You can’t fix what you can’t see

With the convergence of storage and data on the same networking fabrics, it becomes more difficult to diagnose and troubleshoot issues involving the ways in which business information flows. We need to be more wary of how to mix traditional data traffic with SAN and real-time streaming media.

As new technologies get implemented and intertwined in the data center, visibility becomes more challenging.

Take data center network automation as an example. Network automation lets an enterprise do more with fewer resources, which frees up personnel to focus on activities that contribute directly to business growth. But automation means dynamic configuration of network resources, and it requires impeccable network security, audit trails for change management and expert systems for automatic conflict resolution.

As more functions are automated, can you detect the impact on the network? Is your network infrastructure still secure, or have you inadvertently opened potential avenues of breaching by viruses or unauthorised users intent on accessing confidential information? Are the physical network pipes that were provisioned still adequate for the new flows of business information? Can you predict provisioning and resource placement in an automated environment of dynamic configuration? When you add a new technology or resource to the network, do you have visibility into the results? Or are you operating blind, which is a recipe for disaster?

Another important technology to consider is virtualisation, of which network automation is a key component. By increasing utilisation of physical assets along with the ability to share resources with other locations, virtualisation can reduce the pressures for physical space, power and cooling in the data center — all of which leads to lower costs.

Despite the potential of virtualisation, however, there are caveats. Virtualisation intensifies the integration between compute and networking resources, requiring a delicate balance between utilisation and uptime. You need to design your virtualised solutions so that mission-critical applications don’t collide with non-mission-critical applications, and that secure applications are shielded from the impacts of open applications.

Think ahead to maintain business continuity

I suggest making visability a high priority in your data center networking design and planning to ensure you have the right level of visibility to be both proactive and instantaneously reactive.

Visibility means collecting and filtering the right kind of data to allow for trends analysis to find high-risk areas. It also means honing in on and eliminating a source of trouble before it breaks your business continuity.

How do you achieve this kind of visibility? Unfortunately, it’s not as simple as buying some kind of “ultimate data center visibility tool.” Visibility into data center networking depends on your entire network infrastructure design, combining sensors integrated into your network, and the collector of information that centralises it. sFlow is the leading packet sampling standard to help you achieve that level of visibility. HP ProCurve supports sFlow across its wired and wireless networking portfolio, as well as network and security management solution.

By all means, explore the exciting new technologies that will take your data center operations to the next generation. But be wary of potential new failure domains, and make sure your greatest IT asset — your staff — is not one of those failure domains. Keep things simple and work with a partner that provides great visibility into underlying operations. Minimise your chances of problems by making smart network infrastructure planning an integral part of your data center initiatives.

For more information on sFlow and network visibility, visit www.sflow.org.
 
Paul Congdon is Chief Technology Officer of HP ProCurve Networking, as well as an HP Fellow.He is responsible for specifying, architecting and designing ProCurve network infrastructure and software products and he heads ProCurve’s worldwide security strategy. In his 20+ years in the networking industry, he has become widely esteemed as an inventor and leader in driving networking industry standards.