In the last 12 months networking vendors have discovered that they are extremely relevant to the spread of system automation, especially virtualization and cloud computing.  Today’s Cisco Unified Computing announcement has capped a series of strategic developments taking place in the last 12 months and at the live analyst and press conference event where the closing slide promised more to come.

There is more, and Prem Jain was almost baited into talking about more during a Q&A session about inter-cloud movement. Chambers alluded to an interesting potential lunch chat; would have liked to have been there to listen, with or without soufflé.

There has been a surge of significant developments in the evolution of infrastructure 2.0, starting with a groundbreaking and strategic partnership between Cisco and VMware evolving further into Cisco's recent virtualization-enabled data center "invasion", centering on strategic partnerships including Cisco’s entry into the blade server market via Intel’s Nehalem processor. 

Earlier this year Juniper and IBM kicked off a partnership as IBM launched their own dynamic infrastructure initiative.  Cisco and Infoblox have also collaborated on a new branch office solution as they led the Infrastructure 2.0 charge, also with the likes of F5, who has been blogging about I2.0 at DevCentral.

Executives from Cisco, F5, Infoblox, Replicate and VMware are also weighing in on the conversation at the Infrastructure 2.0 blog.

Here Comes the Cloud

Some of the world’s premier networking vendors are aligning with key players and entering strategic peripheral markets to get more footprint in the IT charge to a more powerful, dynamic infrastructure.  What Cisco calls Unified Computing IBM may call Dynamic Infrastructure; they both could easily be subsets of the vendor-neutral Infrastructure 2.0 meme.

Automation is key to the proper evolution of virtualization and cloud.  Here is an interesting blog perspective on IBMs very recent efforts to make a case for greater automation in IT:

To put it in plain American, we want to do for all the things we build to distribute power, products, and people what we did to bookkeeping: automate the living daylights out of it. And, if I read my IBM brochureware right, there is a hell of a lot of money at stake in doing this. – Timothy Prickett Morgan, The Four Hundred

While some say that Cisco’s move into blade servers is about a new market, it is likely more about synergy, or the emerging strategic importance of the hypervisor as a network appliance as networks and systems and endpoints evolve.  This potential is more than a larger market per se; it is a strategic inflection point from which Cisco can accrue other benefits as data centers move to virtualize larger portions of their production environments and contemplate or explore the cloud.

The Net Effect of Today’s Announcement

In effect, a few powerful players with deep subject matter expertise could create a superior level of interoperability, scale and automation and elbow out the weaker players still wringing profits out of the complex, manually-managed container game.  Indeed, virtualization may have marked the beginning of the end of that game and the clerk infrastructure culture described by Urquhart.

These collaborations and entrances are setting the stage for a new age of networking, driven by unprecedented levels of network automation and intelligence.  In this new era applications, networks and endpoints will be in a kind of constant contact regardless of location, enabling location-independent security, traffic and application policies, among others.

Think standards-based, meshes of blade servers distributed in clouds of processors linked by dynamic infrastructure (infrastructure 2.0) pulsing as loads, demands and user requirements shift; saving massive amounts of energy and human resources once trapped in mazes of idle power and network cables, screwdriver-constrained racks and procedural checklists updated manually by overworked network pros doing their best to keep up with heightened velocities of change. 

I think today’s conference acknowledges that Cisco cannot do it all, and would benefit by a fresher, kinder approach to partnership; one that includes back to back sound bites from VMware and Microsoft executives.

This age will move the network from reactive manual configuration to unprecedented levels of dynamic policy and automation capabilities.  It will drive us toward new capabilities including IF-MAP; and will require cooperation among many players, including VMware and Microsoft.

This is where I think Cisco and other networking players (including F5 Networks, Juniper and others) are going.  This is where the network industry redefines itself by declaring war on today’s static infrastructure shortcomings, from outdated legacy tools and processes (including reliance upon spreadsheets and DNS freeware) to layers of waste and delays and unnecessary outages.

If you’re not one to get too caught up in vendor vision announcements about future possibilities you can read about Grainger's success with core network service automation thanks to the special insert in the March 16 issue of Network World. 

Companies like Grainger have discovered the power of automation to reduce network costs and boost availability.  While many enterprises have begun the virtualization process within manually configured VLAN containers, companies who understand the significance of network automation to the IT business case-will probably be the leaders driving critical innovations in virtualization and cloud computing.

While Google and Amazon wow the markets with cloud computing announcements and positioning exercises, enterprises are driving their network costs lower while making them more strategic.  Those who drive expenses out of their infrastructure while boosting availability will likely be the last to sign up for multi-tenant public clouds; the move to public cloud simply won’t be feasible from an economic or security risk standpoint for enterprises already adopting best practice architectures.

The cloud players still need to work through security/compliance, lock-in and availability/scale and storage issues.  They will likely appeal to those with little interest in cloud security while Cisco and VMware go after the Fortune 500 with the energy, economy and flexibility payoffs (part of “operational excellence” according to Chambers).

That’s why I think Cisco’s Urquhart was right when he recently suggested that cloud computing markets would probably develop along the lines of the storage and transportation industries, with a multitude of “flavors” for a multitude of specialized needs.  Those service providers who understand this may become the new IT leaders of tomorrow; yet enterprises that get it may never need the public cloud.  They will simply be keeping up with cloud-influenced user demands.

This will be a topic at the upcoming Interop panel on “Where Can Things go Wrong?” at The Enterprise Cloud Summit.

Cloud is a great opportunity for service providers who need more than speeds and feeds and SLAs to differentiate themselves.  The question is: do they truly understand how important the network is to the pulse of loads, automation and movement required to make the public cloud a realistic business proposition? 

The network is the foundation for many types of cloud deployments; it makes the transport and flexibility across a data center or even continent possible.  Security and compliance are big issues, yet many cloud providers have preferred not to discuss the sordid details of what they are doing.  One has to wonder just how many enterprises are using cloud and yet don’t know it.

The jury is still out as to the scalability and business case for public cloud, as the great Google and Amazon message factories churn out memes ahead of profits and insist for the most part on secrecy when it comes to critical topics like cloud security.  As long as they choose baffle marketing mystery over technology brilliance they will lose to the Cisco’s and Microsoft’s of IT who understand enterprise requirements versus secret sauce slide ware.

That’s why I found today’s Cisco event refreshing.  Company execs explained where they had been and where they were going and how they planned to get there.  There were clearly issues left to resolve and partners who likely would help them.

Google and Amazon should take a lesson from Cisco.  They will need to be more transparent about security, compliance and other issues to become realistic forces in enterprise IT.  That being said, for SMB’s less concerned with security, compliance and availability, the multi-tenant public cloud may be an acceptable compromise for a variety of pre-enterprise applications. 

It is clear from the last 12 months that IT is entering a new era, enabled by the evolution of infrastructure 2.0.  That new era promises more choices than ever enabled by an even more powerful, strategic network. 

One of the first layers of this evolution is the automation of dull, high risk manual network tasks, including DNS/DHCP and IP address management and other network configuration kludge works.  This could be followed by unprecedented automation and resource savings enabled by new storage, I/O and infrastructure breakthroughs unleashing VMotion and the business case for emerging cloud management solutions and partnerships.

It is about time that the network vendors embrace the kind of automation that they have been delivering to users.  Perhaps they were all just waiting for the hypervisor.

I’m a senior director at Infoblox.  You can follow my ramblings in real-time, including coverage of cloud computing at www.twitter.com/archimedius.