A while ago, Greg posted a picture of a fancy, self-cleaning London public convenience. This is a perfect metaphor for what I am about to rant about. This is the height of ablutionary innovation. Self-cleaning and deodorizing. Pleasant and convenient. Presenting the user with the best possible excretory client experience in the public space. However, without adequate investment in the behind-the-scenes infrastructure, all this innovation is just so much (ahem) crap.
For as long as I can remember, I have been attending conferences and seminars where the buzzword is “innovation”. Innovation to increase productivity, to improve user experience, to improve the business bottom line. All very interesting, but invariably upon arrival, the innovation seems to be outside the network space. Let’s face the reality here. Networking (and most other IT infrastructure) is plumbing. And this is appropriate, given that much of what passes for “innovation” is crap, as far as I can tell. To be more specific, most of what is publicized as innovation is, in fact, marketing. “Real” innovation in the IT infrastructure space, at least that which has a tangible impact on us as networking professionals is driven by the need to find solutions to real-world problems caused by the marketing “bells and whistles” innovation.
What do I mean by that? Take for example a lot of the “innovative” platforms I’ve seen come and go over the past few years. In the education space, these have included on-line learning (Blackboard, Moodle, etc), mobile devices, interactive voice and video. Each in their turn have been called “innovative”. But all of the innovation is at the front end. Take the on-line learning management systems. When you break it down, to my inexpert eye, what do these systems comprise? Surprise, surprise, they are a honking great database behind some clever APIs and web presentation. This is not innovative. Support for mobile devices – where is the innovation here? Really, someone tell me. Isn’t it just an exercise in reformatting for a small screen? (Apologies to everyone I’ve offended in the past couple of paragraphs).
For us as network engineers, the innovation we need is in the plumbing. It needs to be real, and address issues or problems. Invariably it is driven by the unintended consequences of what I call the front-end innovation. But while the money and the prestige within the organization goes to the front-end innovators, we infrastructure people need to wait for the crumbs to filter down to us. Want multiple nodes to increase uptime? Fine, the industry develops, and we need to buy, load balancers. Your Moodle instance needs to replicate that obscenely bloated back-end database? More bandwidth, please. The Layer 2 domain needs to grow beyond all reasonable expectations? We need things like OTV, QFabric and FabricPath. Is all this hardware hard to manage? Virtualization can help. Need to support lots more mobile devices? 802.11n and more access points everywhere! So networking vendors are out there innovating, but the audience is small and the traction low when it compares to the front-end, public-facing innovation.
How many times have you been told that such-and-such a team is about to implement a insert innovative new thingamabob or doodad here only to discover that the budget doesn’t include the required ten-gig links to their new blade enclosure or that this will break your load balancer but there aren’t any funds for solving that? Or the new innovation is implemented before the infrastructure impact is even assessed, leading to poor outcomes all round?
This is a fundamental problem we have as packet plumbers. Bells-and-whistles and front-end innovation is all well and good (I grudgingly admit), but without the concomitant innovation and investment in infrastructure, these things die on the vine. Too often, infrastructure is the afterthought in the innovation equation at the management level, and we need to fight hard to bring it to the front-of-mind of those controlling strategy and budget. You have to get the infrastructure right, or that new innovative app will get stuck in the S-bend.