Network equipment vendors today are suspended –sometimes in disbelief, sometimes in denial- in a choice between innovation and irrelevance. The world around them has changed dramatically from the first Interop, from the powerful supply chains that have driven time and expense out of dozens of industries to the amazing rise of ecommerce in retail and banking.
Dan Lynch (one of the founding fathers of the network equipment industry) quipped to us after our Future in Review panel last May that networks have indeed fallen behind the world that they helped to create. (You can watch the FIRE infrastructure 2.0 panel and come to your own conclusions).
The core of the network hasn’t changed materially in decades. That’s a critical consideration now that virtualization has triggered the automation of systems in production environments.
Virtualization deployed within today’s networks is contained/constrained by issues related to capacity, integrity and orchestration. Whoever breaks those barriers first will have a strategic advantage over the “denialistas”.
These problems may not be visible to those who work within these networks or to those who are simply out of touch with the growing operational malaise stemming from outdated tools and practices overlaid on increasingly complex networks.
Network equipment vendors need to align their products and visions with the inevitability of virtualization (not today’s virtualization-lite pockets but rather dynamic and dispersed multi-site meshes of processing power) and the longer term likelihood of cloud computing. That will set the infrastructure 2.0 era in full motion, generating powerful new business cases for network gear by enabling network automation.
Out there is the VMware for the network, driving value by similarly automating outdated network processes. Yet IT automation will also require network automation. The network will become a strategic part of the solution.
Then How Did We Get Here?
As the pendulum has swung between various initiatives more pockets of expertise and specialization were created, more certifications, more configurations, more gear, etc; and network teams were charged with creating ever more scripts between ever more devices. As each vendor created a new release addressing customer pain points, new scripts were often needed.
Perhaps that’s why network and security pros could make great Jenga® players. They live every day adding, removing and scripting (while the world around them embraces automation). That’s also why many networking and security pros often resist innovation until a threshold of pain is crossed; that is, when “new” is less painful than “old.”
If you’ve played Jenga you know that it’s more fun to watch others (at least for me with my hands) carefully change the stack. The games get progressively more difficult as changes are made to the stack.
Virtualization has converted some of these “Jenga blocks” into elastic, flexible cores in some data centers -that are still static on the outside (virtual networks or VLANs)- yet the network game hasn’t changed. Ops pros have mouse clicks (automation) while the network teams are still suspended in time-consuming processes designed to counterbalance the risks of manual labor.
That’s opex spend creating the need for more opex spend. Sounds remarkably similar to the case for the virtualization of servers, doesn’t it?
When virtualization crossed into production environments it was the first shot in the coming IT automation revolution. Whether cloud computing is real or whether it comes in three or five years (or whether yet another vendor redefines its existing services as cloud), the network industry has been served notice: VMware, Citrix and even Microsoft have proven how easy and compelling system automation is compared to the hours and even days it takes to simply move a server the old fashioned way.
Automation is powerful and highly contagious as legions of “middlemen” have learned since the 80s and 90s; and those who embrace it win. Those who cling to manual processes are increasingly marginalized by decreasing economics. The larger the manually managed network the higher the per unit operating cost and the higher the outage risk. Opex begets opex.