So I'm tired of going over OSPF LSA types for the fifth time, so it is time for a rambling, long-winded rant.
A long time ago, I came to the conclusion that I was a plumber, and that networking is plumbing. Not because much of what traverses the world's networks is crap, but rather that the network is critical infrastructure in most modern enterprises. And like plumbing, underinvestment in maintenance and upgrades will eventually mean that the poo will hit the rotor blades. The plumber is only appreciated when something breaks.
Of course, in the case of the network engineer, they seem to get the blame for any problem - theirs or not - and only grudging thanks when a problem is resolved. That aside, it seems that many network administrators do not appreciate that plumbing is their primary goal. Making sure the pipes don't get clogged and that lines don't spring leaks. And here I am not just talking about firefighting; springing into action when a problem is reported. Prevention is the key, but prevention requires knowledge. In my opinion, not enough attention is paid to network monitoring, baselining, and knowing where your vulnerabilities lie.
When someone calls up and reports a network problem, how can you assess the nature of the problem, or indeed whether there is a problem without knowing what normal behaviour is? Or for that matter knowing where the problem may lie. It has always been my goal to be aware of a problem before the phone rings to say there is an issue. Some would say that that leads to obsession and paranoia, and I am not in a position to argue. But being able to do so and to pre-emptively address issues is key to making sure the network provides a robust reliable service to the enterprise.
From the organizational perspective, though, the plumbing is not seen as essential to business until the CEO can't take a dump in the executive washroom. Similarly, the network is not seen as a critical business component until it goes away and starts costing the organization money. So what results is an underinvestment in infrastructure in favor of high-profile IT projects. And by high-profile, I don't just mean visible to the executive, the public or relevant to the bottom-line. I mean high profile to the users. The workgroup switches might be five years old with failing fans and power supplies, and unable to provide new features like security or PoE, but the desktop refresh program presents shiny new PCs to each user every 24 months without fail.
The fundamental nature of networking (and server administration to that end) as a backroom service which doesn't interface directly with the client is detrimental to the recognition of the need for infrastructure, and in fact contributes to its own demise. organizationally. In a client-focused organization, when a user has a problem, they have a cheery help desk to call or a friendly client service officer to come on site. In my experience, these cheery client service people are the ones ready to stab the backroom boys in the back. "It is a network problem". "It is a server problem". "Just fill out the survey to say how wonderful the client service is and by the way here is your shiny new laptop".
And as the network and server environments are seen as a hindrance to client support, they are a prime candidate for either outsourcing, or in today's world, being moved into "the cloud". Who needs server admins or storage admins if we can shove it all into the Amazon cloud? But we still need the friendly tech who can drop off your new PC and give the vendor a call when the monitor stops working.
Just as we would like to have a world where we don't need to engage with the nasty world of moving effluent from one place to another, it seems that enterprises don't want to deal with the network infrastructure or infrastructure of any kind. The connection from the desktop to the service should be transparent, and because of that, it becomes invisible. Transparent it should be. But that requires effort and investment. It can't be invisible. Perhaps a little less client-focus and a little more mission-focus should be the thing?