Even if its unified computing products don't take off, it will push the market where it needs to go.
Why would anyone reinvent the wheel? Well, because then it's your wheel.
Cisco wasn't impressed by the hard-to-manage, virtualization-ignorant servers that Hewlett-Packard, IBM, and Dell make, so it decided to make its own. And along with that new hardware platform, it's teaming with some of the best and brightest data center suppliers to make a system that promises to be the most virtualization-friendly to date.
In doing that, it's betting that it can master the art of blade server manufacturing faster than those rival blade makers can move onto its switching turf. If Cisco's right, the dominant networking vendor will have cracked a major new market. And even if it's wrong, it will have given the blade server market a swift kick that's likely to challenge blade designs and leave businesses with more options.
The networking vendor set up this competition with the launch of its Unified Computing System plan, in which Cisco will sell a new series of blade servers that includes storage capabilities along with virtualization and server management software. Partners include BMC Software, EMC, Intel, Microsoft, and VMware.
Blades as a design are less than a decade old and make up 20% of the server market, says IDC, and they're one of the most profitable server segments. Cisco's idea is that revamped blades can be combined with converged network components, then linked to virtualized network and storage resources under a single management interface.
Companies are likely to move in this direction, since it promises to drive costs out of the data center by prompting more-efficient utilization of devices at every level. It also would move the benefits of virtualization beyond server consolidation into overall management of sets of pooled resources--servers, networking, and storage--again reducing IT operating costs. Cisco says savings could hit 35%.
To accomplish these goals, Cisco must bring a new type of blade to market that relies on a 10 Gigabit Ethernet fabric to feed I/O to both storage and networking devices.
That idea isn't completely new. Since January, HP has offered its BladeSystem with Virtual Connect Flex-10, which divides overall I/O between the enterprise network and iSCSI storage. HP has an interface, Insight Dynamics VSE, to manage the two. Gary Thome, director of strategy for HP's blade and infrastructure software unit, paints Cisco as the "plumber" of network pipes and HP as the general contractor building the house. "We don't see the data center as a network with servers hanging off the end," he says.
Several blade manufacturers, including HP and IBM, already have or soon will have the capacity to combine network traffic and storage data as a single stream coming off the blade to networking devices on the blade chassis. Cisco "incrementally goes a little further" in driving that convergence, says Burton Group analyst Nik Simpson.
But an incremental step may give Cisco a foothold, since it's the first major entrant into the blade market since Dell in 2001. Still, it will have to win over a new layer of data center managers without the incumbent advantage it has in networking, which will require continually refining the design and manufacturing of blades, something that Dell, HP, IBM, and Sun have worked on for years. "Cisco has never moved into a market this size with so many established incumbents," Simpson says.
Cisco knows who's in its way and how its strategy can disrupt networking sales partnerships with HP and IBM. "We focus on market transitions, not competitors," Cisco CEO John Chambers says. "You've got to play the hand this market has dealt you."
BMC Is Pivotal
To HP and IBM: Bring it
Cisco has assigned a key role in the process to BMC, the systems management vendor, to deliver a combined management tool to supervise the different virtualized resources. Providing one management interface for servers, networking, and storage has been a stumbling block. The storage industry hasn't come up with a single interface for storage, because of competing vendors and ineffectual standards.
This time, virtualization vendors Microsoft and VMware--both partners in the Cisco project--may have the heft to overcome those barriers. Each can lay down a storage API for virtualized systems and insist storage vendors link their devices to servers through it, Simpson says. Such a unified storage pool is needed to realize the benefits of virtualization.
To move virtual machines from physical server to physical server, the data center needs a single cluster file system underneath them. VMware gets around the fractious industry by using its own cluster file system. It's promising a storage API in its vCloud initiative that will let any storage vendor adopting the API participate.
Beyond storage, there's still the challenge of delivering a single management interface that cuts across servers, storage, and networking. Cisco has vested BMC with delivering a tool to orchestrate communications on a 10 Gigabit Ethernet fabric. It will understand how to allocate work across pooled networking, processing, and storage devices, monitoring and adjusting as loads change.
Current systems require managers to pre-allocate resources, and the tendency is to overallocate. Since future demand can't be completely predicted, server admins guard against failure by piling on more CPU cycles and more memory and storage than is needed. To balance workloads well enough that managers don't feel the need to overprovision, Cisco needs a new management tool that can view each virtualized resource and manage them as part of a single system.
BMC's flagship product, Patrol, was among the earliest to manage cross- vendor virtualization. Now it will try to graft on cross-resource virtualization. "The pain point," says BMC CEO Bob Beauchamp, "is where virtual machines and the virtualized network meet up with storage."
Cisco and its partners will truly simplify data center operations if Cisco indeed produces standard Intel-based blades with a unified management tool that assigns them VMs, allocating and reallocating their resources based on need, and it's able to move them around physical servers atop a virtualized storage system.
The manager of the virtualized data center "shouldn't have to worry how many network interface cards are on the server or how much memory is available," says Beauchamp. Indeed, the Cisco blade design is aimed at making such configuration problems an issue of the past.
We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.