VMware launched NSX, its Network Virtualization platform at VMworld last week. In his keynote, VMware CEO Pat Gelsinger portrayed Network Virtualization as a very natural extension to what VMware accomplished in Server Virtualization. However market fundamentals and early drivers for Server Virtualization are not quite the same as Network Virtualization. Hence any comparison and contrast between the two should be understood and weighed on in their respective contexts.
The drive for Server Virtualization fundamentally was an attempt to address the growing gulf between faster rate of technology advancement in server space relative to customer ability to utilize the excess capacity. It was a trend that was driven by the focus towards gaining efficiency in an era where cost was becoming important. Over nearly a decade now Server Virtualization has accomplished this goal of better utilization of assets: And server utilization levels have increased by a factor of 4 over the years.
Networks in the data centers today however do not suffer from this excess capacity problem. If any, the problem is the reverse – user demand for networks capacity continues to outpace what is currently available. As long as there remains a growing gulf between user expectations for capacity relative to technology advancement there will remain opportunity for vendors to innovate in this space. In other words unlike the server world, network virtualization does not shift the value away from the underlying infrastructure.
Server Virtualization is transforming IT by providing greater business agility. Goal of Network Virtualization should be to bring similar business agility for the network. However, this goal need not require complete decoupling of the virtual network from underlying physical network as some vendors may lead you to believe. Any goal of gaining agility by completely decoupling physical and virtual network can only be done with some confidence, by significant under-provisioning of the physical network. For if the bandwidth is plenty the overlays have less dependency on understanding or integrating with the underlying infrastructure. This shortsighted approach, which focuses on business agility, but ignores business assurance, will increase the network capital expenditure and operating expense spend over time. Note that even in the server world where compute efficiency was attained, the benefit did not come at any capex or opex savings. Capex savings attained on server hardware was offset by increased cost of virtualization software. And we have seen opex continues to increase over the last decade.
As IT increasingly begins to take on a service centric view, more intelligence will be needed at the edge – physical or virtual edge. Cisco’s launch of Dynamic Fabric Automation (DFA) last July, address this view of an optimized fabric infrastructure with a more intelligent network edge that can enable any network anywhere, supporting transparent mobility for physical servers and virtual machines. Application Centric Infrastructure (ACI) takes this a step further by enabling application-driven policy automation, management and visibility of physical and virtual networks. They however also integrate the physical and the virtual network for an agile service delivery that also assures full lifecycle user experience.
True Innovators are not easy to find. In fact, people who build next-generation innovations and succeed are certifiably rare. Repeat successes with these innovations are rarer still. Hatricks are legendary. Anything beyond is best relegated to a rarefied stratospheric atmosphere and dismissed as fiction.
Fortunately for us at Cisco, one need not look too far for such innovators. The team of Mario Mazzola, Prem Jain, Luca Cafiero and Soni Jiandani (affectionately called by some in the industry as the MPLS team) has both the reputation as well as the track-record of not just building world-class innovations, but for successfully converting them into multi-billion dollar global businesses within the Cisco fold. So, when they talk about networking, a subject which they know a thing or two about, most people find it worth their while to listen up.
Today marks an important milestone for one of our most strategic data center products and the foundation of virtual networking portfolio. Five years ago, the Nexus 1000V virtual switch was the pioneer in the virtual networking market with its launch at VMworld in 2008. Since then it has been adopted by over 8000 customers and continues to grow on other platforms, such as Microsoft Hyper-V, and soon Linux/KVM. Today, Nexus 1000V represents the largest software controller-based networking solution (aka, Software Defined Networking or SDN) in the industry.
We continue to add hundreds of paying customers every quarter, in spite of offering a fully featured no-cost essential edition. The interest in the virtual networking space also continues to increase ever since the SDN trend started. There are also plenty of FUD or rumors being spread about the Cisco’s virtual networking solution. On this 5th year anniversary, let’s do some myth busting focused on Nexus 1000V based solutions. Read More »
As the long awaited innovation in the networking space moves out of hype cycle and market interest in software defined networking (SDN) steadily rises, Cisco has been actively involved in these emerging trends, working with standards bodies and listening to the requirements of our global customer base.
As we continue to make our networks more open, programmable, and application aware, we believe we have the industry’s most comprehensive portfolio to help lead this change in an evolutionary manner.
During engagements with our global customer base, we have heard many claims about SDN. I will address these claims from a customer support point of view.
• SDN is about virtualizing the network. It is about migrating from a static, complex physical network platform to a dynamic simplified software enabled virtual platform
• SDN is about commoditizing network hardware because software can provide all functionalities in a centralized, limitless fashion.
• SDN is about reducing TCO and increasing agility. It is about reducing cost (OPEX) through simplification, virtualization, and automation, but also accelerating innovative business services for growth.
Let’s look at the implications of these.
#1: SDN is about virtualizing the network.
This is true and there are benefits. But before you think about virtualizing the network using “Software Defined Network” or “Software Defined Data Center”, let’s recap some core requirements any IT organization needs to take into account:
• Do you have an understanding of your business application environment?
• What are the key interdependencies between your application strategy and your infrastructure strategy?
• What implications would virtualizing the network layer have on your SLA’s? Are there performance penalties associated to your business if you don’t meet them?
If you have not explored these questions in details, then consider developing an “application to infrastructure” blueprint that is aligned to your business strategy. Leverage SDN as a crucial technology building block that can accelerate this process and provide solutions to any gaps identified hence simplifying your path towards network virtualization.
#2: SDN is about commoditizing network hardware because software can now provide all network functionalities.
This is not a reality today. The evolution of PaaS/SaaS and application providers support the fact that software is not “limitless.” The need for network intelligence, scale, performance, and security are still top priorities of most IT infrastructure. SDN does not promise to eliminate the importance hardware has, but simply illustrates the possibility of moving the decision intelligence from the hardware to software. What about speed and performance requirements on a software controller? Can it scale and grow as fast as the business (traffic) needs? How about the hardware that the controller software is running on — can it react fast enough to the ever-growing computing and storage demands? Can your “software only” infrastructure grow dynamically and as fast as your business application needs grow? Each organization needs to consider the implications of transferring the risk and complexity from infrastructure deployment to software (controller and agents) development.
#3: SDN is about reducing TCO and increasing Speed to Market for innovative business services.
Total Cost of Ownership calculations include both CAPEX and OPEX.In an “SDN” world, CAPEX = hardware cost + software cost which includes both development and maintenance, whether you choose to develop in-house (i.e. hiring new skills or transform your existing staff) or through a third-party software developer (i.e. licensing and upgrade costs). Is your operating model changing fast enough to utilize the benefit of SDN’s *simplified* management and operation? Have you broken down IT siloes between Security, Compute, Storage, and Operations? Is your organization ready to shift from managing infrastructure to managing software and changing the IT operation structure? Do you have the necessary tools and process to capture the rich data an “SDN” architecture now provides and turn them into new services for creating new revenue streams? In other words, is everything going to get more complicated before it can be “simplified”? Looking back to the server virtualization transition, very few IT organizations, if any, can claim that they realized the projected operating ROI within the first few years.
Now, let’s take a look at Cisco Open network Environment (ONE).
CiscoONE is more than just SDN. It offers a solution set that provides:
• A softwarecontroller that is centralized and separated from the local data planes
• Network programmability
• Network virtual overlays
We see the move to programmability and network virtualization as an evolution, not a revolution. CiscoONE creates incremental functionality that can be layered on existing infrastructure to deliver new functionality and provide SDN capabilities on top of both traditional technologies and modern business application needs. This enables you to continue leveraging value from the IT investments you’ve already made. We are working to help customers extract more value from the network by offering a broad network API, rich features, and analytics. The core value of the Cisco solution is an “Application Centric Infrastructure”, compare to the generic “Software Defined Network” term. Both software and hardware are a means for providing services business applications. And by making applications the center of everything, we take the broadest view to deliver openness, programmability and abstraction across multiple layers, to the line of business owners.
As my colleague Stephen Speirs pointed out in his blog, Services is the missing S in SDN, I would also say Strategy is the starting point for SDN. Why Strategy?
With the right strategy, you can plan, build and manage an open, programmable, “virtualized” network that reduces your OPEX and delivers the business outcomes you need with the minimum level of risk.
– Customers are at various stages of SDN adoption, as with any new technology or network paradigm. We’ve heard a lot of questions from them: How do I build a business case for SDN? How do I validate the ROI? How do I manage SDN devices? How would SDN change my operations model? What new security vulnerabilities and regulatory issues will I have? How do I build the API applications that are needed for my use case? What do I do when something goes wrong?
– Few customers have a clear understanding of their application profiles. Without a clear view of your application profiles, there are risks to deploying SDN.
– Cisco Services can help you through the SDN journey starting from identify the right strategy to execution so that your organization can transform your business agenda to maximize business value and minimize risks.
Services help you address the areas of What, Why and How
The Cisco Services team is well-positioned to lead this transition for customers. Our work with enterprises, services providers and public sector organizations over past 20+ years has provided us with unique network insights and implementation experience. Cisco Services offer consulting, professional and technical services via strategy, assessment/planning (Why), design and development (What), deployment, validation and operations services (How). We have the experience to help you adopt open, programmable or virtualized networks based on where you are today and where you need to be in order to harness Network Intelligence through deep programmatic access to your networking platforms.
It is amazing how the data centre world has changed in the last few years. A Data Centre used to be a collection of network elements to interconnect static servers (and their associated storage), with traffic patterns that were highly predictable and mostly north-south. Cloud and virtualization have changed all of this: a data centre is now a collection of compute and storage resources which can be securely sliced up into virtual networks and placed anywhere according to real time needs, interconnected by a fabric. The virtualization of servers, network services such as firewalls and load balancers, and even network devices such as switches and routers, has created a very dynamic landscape in terms of how fast you could configure a virtual network, in a way where location shouldn’t really matter, and where compute and storage resources can be added on the fly, based on demand. Multi-tenant Data Centres, such as the one to deploy Virtual Private Clouds, need to support 10000’s of these virtual networks. And every one of these virtual networks needs a lot of different service instances to stitch together the virtual network across virtual servers, virtual switches, virtual firewalls, virtual load-balancers, and virtual routers. Traffic patterns have shifted to East-West, because of the new applications which spread processing across many hosts, and because of the ‘location freedom’ that virtualization allows. Network infrastructure needs to be cost-effective to handle all this traffic, while the increased lookup-table size caused by the any to any traffic patterns often led to increased cost. Read More »