Cisco, vCider, and the Risks for Software Defined Networking
Fall is here, and everyone is talking about cider.
But while this year's apple pressings are reported to be particularly good, that's not the kind of cider they're talking about in the IT shop. In this case, it's CIDR, or classless inter-domain routing, and more particularly, Cisco's purchase of a small company called vCider. The firm specializes in a unique form of virtual network overlay technology designed to allow enterprises to provision and maintain distributed network architectures on the cloud with all the performance and security features that are currently available in the data center.
The plan is to fold vCider into Cisco's Cloud Computing division where it will support the Cisco version of the OpenStack protocol as part of the company's Open Network Environment (ONE) portfolio. In essence, vCider provides the multi-tenant distributed virtual network controller that will tie in with OpenStack plug-ins and development tools that Cisco gained from the recent Quantum acquisition, giving developers a handy way to streamline both virtual and physical network connections between local and cloud-based resources.
Naturally, this has led to widespread intrigue and speculation as to Cisco's plans for the software-defined network and how those plans impact the vast array of partners/rivals all vying for dominance in what is expected to be the next major transition for enterprise data environments. On the one hand, Cisco has embraced OpenStack and SDN even though it poses a very real risk of undermining its own networking hardware sales − the company's bread and butter. Cisco counters this by saying it can provide a value proposition by integrating management of both physical and virtual resources to drive greater performance and efficiency than if an SDN layer were floating atop commodity hardware, and the vCider acquisition seems to support this contention.
How, then, do we square this with Cisco's continued cooperation with EMC and VMware, which are only marginally involved in networking hardware and are eager to foster pure SDN through acquisitions like Nicira. Cisco, if you recall, was in the running for Nicira as well, but apparently balked at the $1 billion-plus asking price, and now is facing the very real possibility that a fully independent virtual networking layer could take hold in the enterprise due to VMware/EMC's dominance in both virtual and storage technologies.
Again, enter vCider. With the advantage of a unified physical/virtual network environment, Cisco stands a good chance of blunting Nicira's, and by extension VMware/EMC's, influence over the SDN market. And that gives Cisco a better change of managing what appears to be the inevitable shift from hardware-centric network infrastructure to virtual/logical environments.
For the moment, at least, all appears to be hunky-dory for this SDN triumvirate of Cisco, VMware and EMC, not to mention other top vendors like HP and IBM. As long as everyone supports OpenStack, there is at least the appearance of one big happy family. And as long as rivals like CloudStack and Eucalyptus pose a threat to SDN dominance, there's more to be gained from a unified OpenStack environment than a disjointed one.
But as we saw in ancient Rome, pacts of this sort do not last forever. Real world events have a way of driving wedges between partners, and cold, hard business takes precedence over cooperation and, sometimes, even consumer preferences.
I'm not saying this will definitely happen to the OpenStack party, but with A-listers like VMware and Cisco each backing separate technologies in support of SDN, the exit doors are clearly marked.