VMware's purchase of Nicira addresses a longtime deficiency in the company's private cloud offerings and network strategy.
Nicira came out of stealth mode earlier this year with a list of large cloud providers and other large Internet companies as customers. Its software-only approach lines up well with VMware's approach to virtualization and is an important addition to the vCloud. While VMware gets substantial talent with Nicira, it'll have some work to do to integrate the product and will have to look out for its relationship with Cisco.
Nicira is a startup focused on using software-defined networking methods to manage software switches in cloud networks. Primary products include a proprietary fork of the Open vSwitch project, which provides software switching in open source hypervisors and a management platform to automate the configuration and management. Nicira founders and management are part of the OpenFlow consortium and use OpenFlow concepts to drive their technology.
Nicira has a "team of all-stars" for networking. In a real sense, VMware is acquiring not only a real product that works for cloud networking, but also a team of people who can identify technology, determine needs and then develop products to fill the need. Not inconsequentially, people like Nicera founder Nick McKeown and others have been at the root of many networking advances in the last decade. The talent acquisition must be a significant attraction for VMware, as it got not only people who can write code, but also people who can change the networking market with new standards, ideas and technologies.
What This Means for VMware
VMware has failed completely to deliver a network strategy for its customers. The company's vDS/vSwitch isn't a networking product, just a NIC enhancement. Previous networking attempts like vCloud Director Networking Infrastructure were badly received, and VMware currently relies on Cisco to provide networking via the Nexus 1000V software switch.
VMware's relationship with Cisco doesn't bode well for other vendor partners such as HP, Brocade and Dell, which all would like better network integration with VMware. A Nicira-derived solution would certainly help widen the partner appeal--not everyone loves Cisco.
On the other hand, Nicira products don't appear to work with VMware today. Its switching code is largely based on Open vSwitch, which is more normally a part of OpenStack. As Colin McNamara points out in his blog post, the value of the hypervisor is zero--it's the management platform that delivers value to customers and generates the revenue. VMware might be positioning to extend its vast management portfolio, which includes many software assets from EMC (such as Ionix and Greenplum) to include OpenStack and other Open vSwitch platforms.
What Must Cisco Think of the VMware/Nicira Deal?
It's possible that this puts VMware into competition with Cisco. You could consider that EMC owns about 80% of VMware and could readily move to set up a competitive hardware business with servers and networking equipment to build a one-stop shop that competes with HP, Dell and IBM. Networking and server hardware for a data center platform would be reasonably straightforward to buy from the usual ODMs. Certainly, EMC has manufacturing, distribution and sales teams well positioned to execute on such a strategy. But it's much more likely that Cisco and EMC will redouble their efforts in VCE and sell vBlocks--that business is already running and starting to make some impact. (Probably not enough impact to make shareholders happy, but it's probably too soon to tear it down.)
On the other hand, Cisco has its Insieme spin-in, which is widely rumored to be developing either something like Nicira or, more likely, data center automation and orchestration tools to manage all of the data center infrastructure. While EMC has RSA for security tools, Cisco has firewalls, load balancers and many other service-focused technologies.
Cisco certainly isn't standing still--it will soon complete its next generation of virtual networking features in the Virtual Machine Fabric Extender (VM-FEX) and Nexus product families. These new features will be built on the IEEE 802.1BR and 802.1Qbh (both are bridge port extension standards) and offer a much more dynamic solution within the network hardware layer. Cisco appears to be delivering networking products that use both software switching via the NX1000V and hardware networking via FEX networking. Since the enterprise market is much larger than the public cloud market, it's possible that software products could rapidly wither away when viable hardware-based alternatives exist.
Regardless of how VMware's relationship with Cisco goes, there are two clear messages to be taken from the Nicira purchase. First: Software-defined networking (SDN) has been validated by serious money. SDN will be the major touchstone for the next five years, and the Nicira purchase proves that.
Second: The Nicira purchase is a strong reminder that networking is the platform on which all computing depends. It's crucial to the industry that a network strategy support already-apparent changes in the upper layers. On the other hand, it's not possible for networking to change rapidly because any networking change will have a significant impact. Accordingly, operating system and applications vendors resist the adoption of new networking protocols or are hesitant to take risks with new networking technologies, so the industry has to work much harder to find ways to innovate without changing anything as viewed from higher in the stack. And that's really hard. Nicira found a way--that's why VMware's paying a billion bucks for the company.
Subsea cable alternatives can provide a level of comfort for those concerned about disruptions. Realistically, they will continue to play a small, but critical role in the short term.
Maintaining existing network infrastructures often incurs huge technical debt before newer technologies are adopted over time.
Amid ongoing regional uncertainties, telecom infrastructure in the Middle East has expand to include terrestrial low-latency network infrastructure, which offers resilience and agility in navigating political complexities without relying solely on undersea networks.
