Solution provider takeaway: Solution providers can learn how to address common blade server objections by arming themselves with the right information and good workarounds.
As you know, blade servers can be a real boon in your customers' data centers. Increased server density in a smaller footprint; easier server expansion, maintenance and swap-in; less network closet clutter; and rapid provisioning of new blades -- all these benefits mean a more efficient data center for customers.
But to deliver these benefits, you of course have to get buy-in from customers -- and oftentimes it's a hard sell. You've probably encountered a whole host of objections. But even the strongest ones can be readily countered with education and sound solutions. Here's what you need to know about the objections to blade servers and ways to address them.
Objection #1: Blade servers are a single point of failure.
Blade servers do represent a single point of failure -- but only when they're not implemented correctly. Your customers' networks were been built with an eye toward resiliency and business continuity. This means the data center has multiple power sources coming into it, allowing staff to divide power draws and reducing the possibility of being hobbled by a cut cable or brownout. To avoid problems, server chassis need to be configured to support N+1 power supplies, and power to the servers must be divided between the incoming sources. As you look at blade server options for your customers, it's important to make sure the products you select are able to support a fully configured chassis even when one of the power supplies fails; N+1 power supply configurations keep your customers safe.
Objection #2: Input/output address maintenance is problematic.
Blade server I/O addressing for network and SAN connectivity is not a problem, per se. Addressing is done as the servers are brought online. I/O address maintenance issues arise when the data center contains dozens of blade servers, and the capabilities, connections and storage assigned to a blade must remain available should the device fail or go offline. I/O virtualization architecture solves this problem by providing a centrally managed pool of I/O connections and eliminating the need to physically manage the I/O for each blade. Virtualizing the I/O connectivity lets solution providers deliver new services and maintain existing network services with nominal interruption to existing network performance. The largest blade server providers offer I/O virtualization as part of the management tool suite. But before signing up as a channel partner to the vendor, ask for a demonstration of the I/O virtual addressing capabilities across the full server product line you're looking to support. Some vendors support only a fraction of their blade server devices.
Objection #3: Blade servers put out too much heat.
The total heat output for a blade server is less than the cumulative amount of a comparable number of rack-mountable servers. But heat output does come into play when there are a lot of blade servers in a single chassis. To address this very legitimate concern, you should recommend N+1 fans, especially when implementing a fully configured blade server chassis. This configuration ensures adequate cooling even when a fan needs to be hot-swapped. Another option is to include an environmental monitoring unit within the chassis. APC's NetBotz products, for instance, provide excellent heat and smoke monitoring, plus an option for video surveillance.
Objection #4: Blade servers don't support green strategies.
This argument is just plain inaccurate. A fully loaded blade server requires measurably less power than individual servers -- including power for cooling. By adding an additional fan to the chassis, or a monitoring device to the rack enclosure, companies can build a green strategy by cutting power usage compared with a single-unit configuration. Visit the Blade Computing Community or The Green Grid for independent analysis, product information and white papers on servers and blade server technologies as well as green computing strategies.
Objection #5: Blade servers are expensive.
This is another unfounded argument. In addition to reducing power costs, blade servers reduce real estate costs through a reduced data center footprint. Blade servers also reduce device maintenance and management costs by consolidating individual servers into a single chassis, reducing overall network complexity.
Bottom line: Blade server benefits outweigh drawbacks
Blade servers offer real, tangible benefits to customers needing to implement more than eight to 10 servers. They provide improved data center utilization through a reduced footprint, deployment, management and maintenance convenience, and reduced total power consumption. Like all new technologies, blade server objections can be addressed through careful assessment of customer needs and vendor evaluations.
About the author
Martha Young is co-founder and CEO of Nova Amber LLC, a business consulting company specializing in business process virtualization. She has co-authored three books on virtual business processes: The Case for Virtual Business Processes, The Virtual Worker's Handbook and iExec Enterprise Essentials Companion Guide.