This post originally appeared in the Packet Pushers’ Human Infrastructure Magazine. If you’d like to get Human Infrastructure in your inbox twice a month, or read back issues, you can sign up here. We don’t share your info with anyone else.
Networking technology is moving quickly. Orchestration systems are trying to take over network provisioning. Automation tools are helping make the orchestration vision possible. Overlays are adding a manageable connectivity layer on top of that pesky, difficult-to-manage hardware underneath. Open networking has separated the NOS from the hardware, making it possible to mix and match switches and operating systems. NFV has made easier distributed networking architectures that eschew massive appliances, favoring many small software instances running on x86 instead.
The payoff is financial, with the ability to (arguably) spend less and get more. The payoff is also in efficiency, where routine tasks are automated, and applications are stood up faster. The payoff is in a reduced risk, where a well-designed, distributed network architecture (hard) has a reduced blast radius during inevitable failures.
And yet, most enterprise networks seem to be stuck in the past. Why is this? It’s not the tech. I acknowledge the tech isn’t mature yet, at least not all of it. I’ve bemoaned the immaturity issue multiple times. Even so, I don’t think young tech is at the core of the new technology adoption issue, because an awful lot of it is good enough.
I believe the issue is people.
Now, I said “issue” and not “problem.” I can’t blame people for not getting into new networking tech. It’s who we are.
People are slow to react to new information. This is especially true with new technology, where implementing new hardware or software is far more involved than a casual decision to try something new.
People are resistant to change, and with good reason. Folks fear being displaced by change, where the value they once brought goes out the window due to something new beyond their influence. People will defend their turf, especially engineers with deep knowledge they acquired over years of experience.
Some people are too stressed to handle one more thing, when that one more thing arrives in the form of new technology that might or might not move the needle for them.
Some people are simply ignorant. If you’re a Packet Pushers reader and listener, that’s probably not you. Obviously, you have some intent to keep up with the industry. But some folks don’t have that drive. They aren’t trying, and thus they haven’t even heard of new techniques they could use to improve how networking services are delivered in their shops.
Yet other people try to keep up, but can’t contextualize the information. They don’t know what to do after reading an article about a new networking technology. Is it useful? Is it applicable? Would it have a positive impact on operations? For some, those questions are too hard to answer, and so new technology gets filed away and forgotten about.
Other folks are kept in the dark against their will. Business leaders chronically underfund training budgets and conference attendance. I’m on record as having mixed feelings about most conferences, but I still believe there’s value there. And training is a must! Businesses that don’t invest in their tech staff don’t invest in their future infrastructures. The two ideas go hand in hand.
Of course, business leaders are also risk-averse, and they tend to assign new technology a higher level of risk by default. Because “new” is scary and unproven, so the logic goes. This logic prevails despite the track record of incumbent vendors offering horrid technical support and buggy code releases.
All technology has to be taken on its own merits at this point. The age of a vendor, whether startup or established, is no longer an indicator of the risk their product introduces to a computing environment. Don’t presume that because it worked before, it will work again.
Business leaders are also unwilling to change if there is no clear indication that they need to. Some companies sail calmly along, cranking out steady revenue and a decent profit, with no great drive to change their IT practice, no matter what inefficiencies exist. Some leaders might acknowledge the need for an IT improvement here or there, but there’s not enough pain to actively seek pain relief. What they have is good enough, and reactivity is easier than proactivity.
What About You?
In my experience, hard IT change is often driven by individuals. People who are willing to take a leadership role and grasp new technology, despite the odds against them, can affect change in an IT organization.
It’s driven by people who stand up in a meeting and agree to take on the discovery, the research, the hard task, the complex role, the difficult integration, the questions. The blame. The sniping. The blame thrower aimed at the new switch with the new NOS every time some little thing goes wrong. Those are the people who make IT better, ultimately.
And those are the sorts of folks who are going to make networking better. They’ll do the hard work. They’ll get a handle on the new tech. They’ll make a different recommendation than usual for the next refresh cycle. They’ll take ownership. They’ll take responsibility. They’ll be willing to stand and deliver.
Is that you?