Gregory Ness'  Instablog

My recent podcast on infrastructure 2.0 with John Willis reminded me of some recent banter about my use of the term “Cambrian explosion” to talk about where IT is headed when infrastructure 2.0 becomes a reality. In short, today we have a classic preconditions scenario; that is, when automated systems are connected with automated networks. Synergy is putting it mildly.

 

The nature and extent of how automation is enabled and by whom could have a substantial impact on which vendors have the most leverage with customers and partners. 

 

As the likes of Cisco, EMC, Oracle and VMware make acquisitions and strategic partnerships, the “ownership” of that ugly middle ground between applications, endpoints and networks -now occupied by IT’s “middlemen” and their outdated tools- will translate into competitive advantage.

 

Virtualization unlocked the first wave of capex savings; infrastructure 2.0 will unlock the next wave of opex savings.

 

 

Today virtualization is contained within ever increasing, scattered VLAN empires that some are calling clouds. If anything these clouds are mere vapor pockets trapped in shower stalls.  If you’re new to this discussion you can catch up here.

 

When the barriers that trap VMs convert into intelligent, permeable membranes it will unleash a new era of IT innovation built upon a new level of connectivity intelligence between endpoints, systems and networks. Think of it as a kind of supply chain for connected IT tasks.

 

Today that supply chain is a collection of highly-skilled network pros using decades-old practices, from scripts and spreadsheets to checklists and committees to keep networks available, secure and up-to-date. Kind of takes you back to the days before the network, huh?

 

We saw the rise and impact of just-in-time manufacturing. We’re about to see just-in-time IT, a term recently used by Arista’s Gourlay. Those who embrace automation will win yet again.

 

 

You can read an overview of the Cambrian explosion at Wikipedia, but I’ll pull what I think is the most relevant quote:
 

 

 

Cisco blogger James Urquhart once commented how virtualization will open up a broad range of IT models, similar to the complexity of the shipping industry with warehouses of various sizes, distribution centers and a variety of transport mechanisms (trucks, vans, ships, railroad cars, planes, etc). We’re already flustered with cloud definitions, and most models have yet to fully embrace the network.

 

I think it was Irving Wladawsky-Berger who made the reference at Always On many months ago that resonated with Urquhart’s prescient observation. I linked to a recent piece, yet he has talked about this notion for awhile.

 

Infrastructure 2.0 Young Turk Mark Thiele (from VMware) made a comment recently about how automation should not eliminate complexity but rather hide it from users so that they can focus on critical elements. That hasn’t happened, especially in the world of DNS, DHCP and IP address management for starters.

 

 

So then what happens when IT is freed from the shackles of manual scripting and configuration and change approval committees, and IT pros can focus on policy (security/access, application delivery, network, expense, etc)? Tools evolve leveraging the new infrastructure that enables new potentials, similar to what occurred when desktop OSs evolved (from CLI to GUI for example); users expanded and created markets for more tools for more processes.

 

The dynamic, persistent link between application, network and system becomes the foundation for a cloud OS, which in turn transforms the network equipment industry by reducing opex and justifying new levels of capex, new levels of investment in infrastructure 2.0 gear. That in turn creates new markets for new applications which are not truly network nor system; they are likely virtualization or cloud applications offered by those who saw this coming and planned for the explosion.

 

The worlds of various hardware vendors, from Cisco to Juniper to HP are in collision, creating vast new opportunities and niches within a vastly larger category. Sounds like a potential Cambrian-style development to me. 

Whenever there are technology disruptions there is a period of time when smaller upstart companies and established giants are at a kind of parity. The premium on technology and speed is momentarily higher than the premium on stability, safety and size. Enterprises invest in upstarts (or even larger players) during this period because they can deliver sizable strategic advantages.

 

Today virtualization is creating new demands for yet another disruption. Perhaps the first shots in this revolution were fired last week at an all day working session of networking legends and “Young Turks” at the SRI Infrastructure 2.0 meeting. Time will tell, but the level of interest and discussion was exhilarating and timely. Vint Cerf, Bob Grossman and Dan Lynch did an incredible job setting up the session with the help of SRI.

 

 

My take on why we met: Because of the spread of virtualization the IT industry is already experiencing the effects of an increasing disconnect between system and network automation. Virtualization, for example, increases the rate of change and can create new levels of network complexity (see virtualization-lite) because of the proliferation of static vLANS (virtual network segments).

 

A participant in the SRI I2.0 session told me during the lunch break that vLANS are “ugly, necessary workarounds” (as they continue to spread through production data centers). That’s because the decoupling of application from hardware enabled by virtualization is still tethered by network shortcomings… the final frontier for cloud computing. 

 

As automated systems multiply and move, static networks are increasingly taxed or disconnected by the costs and delays inherent with manual labor, committees and necessary approvals. Yet virtualization will continue, with or without the network team, because of its powerful initial ROI and the ability of the VLAN to restrict movement and allocate security and delivery policies to groups of virtual servers.

 

This kind of virtualization-lite (virtualization by VLAN segmentation) is powerful because it breathes more life into a tired, static network; it also demonstrates the power of automation. It is often the first phase of virtualization being deployed because it minimizes the network effects, at least initially.

 

Virtualization-lite is the “lightning in a bottle” that proved innovation in IT is still alive and can reap truly amazing payoffs. Yet over time it pushes the network envelope and experiences the same challenges it initially helped networks avoid. 

 

Virtualization also represents a cultural shift in expectations and the emergence of early cloud computing economies. Yet more VMs and VLANs mean more complexity at some point; different sized networks will reach that point of obsolescence pain at different times. 

 

 

The initial period of substantial capex payoffs ultimately levels off and new opex burdens emerge as VLAN complexity replaces sheer endpoint complexity. VM sprawl becomes vLAN sprawl and ever increasing OPEX is required to service the existing architecture. 

 

Increasing cost and complexity then sets the stage for the next tech revolution, centered in the network itself, setting in motion new network economies and new cloud computing potentials. Players to watch (including those who aren’t networking equipment vendors) include Cisco, Citrix, VMware, EMC, Oracle, HP, Juniper, Sun, IBM and F5 Networks.  
  

The first companies to drive expense out of the network will have a strategic advantage over those who simply rely on faster, cheaper (capex models) gear as the industry experiences disruptive new realities driven by rising complexity and movement. 

 

Last week Mark Anderson’s paid SNS newsletter highlighted a few comments I made at the 2009 Future in Review Conference about the rising cost of complexity, and then made an interesting comment:

 

 

 

 

 

Yet the next network revolution is about more than reducing complexity and cost, it’s about a new and powerful synergy between automated systems and automated networks. New applications, new economics, new potentials are unleashed by this synergy, transforming the IT industry yet again.

 

Those who view automation as simply labor replacement will again be left behind as automated networks find new missions and sell more gear and make those who run them ever more strategic to larger and larger networks. Companies who master the complexity and drive out costs are the future hope for new careers and innovations. 

 

 

One of the promising new capabilities associated with network automation is stateful VMotion (or stateful live migration). Stateful, frictionless live migration of servers in pursuit of temporary operational advantages is a key factor in the future of cloud computing, especially private enterprise clouds. It will flip the economics of mainframe-era centralization strategies (even those in low cost areas) upside down as networks once again prove the power of the mesh over the isolated container.

 

The ability to shift processing cycles from one facility to another, for example,  based on even temporary changes in expenses, users, taxes or disaster effects, etc. is infinitely more powerful than a one-time static decision made every 10-15 years supplemented by manual shifts and committees. We’ve seen this movie before and we’re about to see it again.

 

 

The emergence of larger and more complex vLAN empires will ultimately create enough pain and complexity to force the evolution of network infrastructure, in the same way that the static network created the massive production virtualization opportunity.   Larger, more complex networks deploying increasing populations of VLANs will be the first to evolve. And thanks to some of my personal heroes in networking the effort may have just begun.

 

 

 

 

Disclosure: Long - VMware

As you evaluate the promise of networking and virtualization stocks the question of how virtualization takes shape could make all the difference. The fortunes of companies like VMware, Cisco, Microsoft, HP and others can fluctuate as their offerings appeal to IT buyers.

 

And the first wave of virtualization projects in many enterprises have had very little to do with the network.  That is not necessarily an encouraging sign. Recently I asked the question of whether or not most of today's networks could even keep up with the new automated systems.

 

That is why a recent blog discussion between Arista's Gourlay, Cisco’s Sultan and myself, followed by a blog from F5’s MacVittie is interesting. It is filled with implications for the future of the network (vendors and careers) and the evolution of IT.

 

The surprising blog conversation is likely an artifact of the success and shape of stage one virtualization (or virtualization-lite). With virtualization-lite system pros convert their legacy servers into images inside commodity blade servers (populated with virtual servers) and IT shops see an immediate economic benefit in the form of increased server utilization, responsiveness/flexibility and energy savings. That is pretty compelling, yet renders the network somewhat irrelevant.

 

In the short term continued growth in servers (now virtual machines [VMs] instead of “tin-wrapped software”) means larger blade servers with higher densities of virtual machines inside them.   Servers can be created virtually and moved (within VLANs) in seconds, much faster than the days it could take with legacy servers and the network processes. 

 

The business case for this first stage is compelling and the initial savings can be used to fund even more virtualization projects. 

 

Imagine an operator at a 40s era switchboard trying to keep up with a moving cellular caller throughout the day?  Imagine the cost of populating enough switchboards to manually track callers?  That is the essence of the challenge that virtualization poses to the network today. More change, more complexity and more endpoints will drive more pain into a network that is manually managed.

 

Mouse click to create a new virtual server then wait two days to assign an IP address, etc.  I don’t think so.

 

This breakdown between system automation and network manual labor is the core of what has made the network largely irrelevant to the virtual machine that has replaced the traditional server. VM economics can be so compelling that the savings from a single project can fuel additional projects and additional savings.

 

That is why network vendors will be rushing to develop the connectivity intelligence required to dynamically link applications, endpoints and networks. There are not enough bodies to throw at that problem.

 

MacVittie’s point in her blog (mentioned earlier) was that VM density was becoming the “new measure of IT efficiency”. If that trend continues the network becomes only tactically relevant to the data center, a development that would please some vendors and be troublesome to others.

 

 

However, there is more than a glimmer of hope in the eyes of networking vendors and professionals, and it’s called VMotion. It’s the ability of that virtual machine to easily move across VLANs (including data centers thousands of miles apart) in pursuit of potentially massive operational savings and scalability.

 

While the VLAN container represents tactical VM flexibility (spinning up and moving servers within a confined area), VMotion represents strategic flexibility or the ability to quickly pursue the lowest cost options for delivering services (electricity, taxation, labor, etc.). 

 

Rather than building a very expensive class A data center, cloud builders can build a global mesh of smaller ones and see both capex and opex (ongoing) savings.   Local changes in rates or regulatory events could be addressed in minutes instead of years. That’s a pretty powerful promise.

 

As Google and Amazon and other cloud providers build massive facilities (typically in areas where electricity and real estate is inexpensive), enterprises could be very cost competitive by building meshes of smaller and less costly facilities continuously optimized for cost based on a global spread of expense options.

 

While networking vendors have been on the sidelines for the spread of virtualization-lite, they will be front and center as VMotion is unleashed. That level of dynamic movement will require new investments in automation and management, especially around addressing. 

 

Like during the emergence of VoIP, the role of the network has been misunderstood (as blogged recently). Large scale VoIP projects grabbed the headlines as enterprises ultimately discovered the network impacts and made additional network investments. Virtualization will be no different as these blade servers continue to be populated with more VMs.

 

When the network vendors solve the VMotion challenge they will create a new dynamic in virtualization that could impact the fortunes of both virtualization and networking players. That work may be already underway, as a team of “network industry legends” convenes in early September at SRI under the title: “Pipes and Clouds: The need for Infrastructure 2.0 (The Intercloud Challenge)”.

 

Networking vendors are also stepping up. Earlier this year Cisco announced their Unified Computing System with advanced networking capabilities.  Startup Arista Networks joined the infrastructure 2.0 race recently with an impressive vEOS announcement:

 

 

Clearly the network industry is waking up to the issue and figuring out how to take virtualization to the next level. At Infoblox we were already planning a webinar with Nemertes, Cisco and VMware executives for late September on virtualization and the future of the network, so the contrast between the blogs and these network industry announcements has set us up for an interesting session. 

 

Disclosure: Long VMW