Partial map of the Internet based on the January 15, 2005, data found on opte.org. Each line is drawn between two nodes, representing two IP addresses. The length of the lines are indicative of the delay between those two nodes. This graph represents less than 30 percent of the Class C networks reachable by the data collection program in early 2005.
There was a time not that long ago when anyone could own and operate a business with little or no computer savvy. If the business got big enough, there might be a need to contract out payroll or accounts receivable to employees or third-party contractors who were PC literate. But you, as owner and operator, could limp along and make a living without a sliver of doped silicon at your command.
That time is long gone. Today, it's difficult to imagine being able to communicate with potential and existing customers without email, social media, or nuts-and-bolts networks like Voice over IP (VoIP). Computers and the networks they comprise are no longer a fad—they're as essential to any business as a telephone was 50 years ago. Unfortunately, they're way less reliable, way more confusing, and way, way more expensive than those neolithic landlines ever were.
And we are now at their tender mercy.
Doesn't it seem like our networks have been struggling lately? Whether it's the Internet as a whole or a specific end-user office set up, our networks behave like traffic in gridlock. We know why road traffic has gotten so bad: It's a combo of not taxing rich people enough and more people on the roads every year. But why are our electronic networks suffering? Below are a few ideas about why this may be happening. But what we really need is some of you network gurus chiming in, in comments. I'll go out on a limb and suggest right up front that money is involved.
The top culprit is what it always is: thievery. Just as a human immune system causes problems if it's too trigger-happy or not trigger-happy enough, security software can be too tweaky, or not tweaky enough. The bottom line is thievery requires countermeasures, which bog down everything from email to page-loading time. This will not get better. In fact, it will probably get continually worse.
Culprit No. 2: the cloud. A cloud is just a server located offsite that serves up applications and data on command. As long as the wires connecting the cloud are working at nominal capacity and the servers themselves are able to keep up with demand, everything is fine. But that's not how the real world actually works. Management types don't like to hear this. They prefer a magical ideal world where demand for data and operations doesn't show up in spurts—some correlated to the time of day, others completely chaotic and therefore unpredictable. That reality means that the network has to be capable of running at peak capacity all the time, which costs more money.
Culprit No. 3: the multitasking, overworked employee. Networks and PCs have created a workplace where one person can do jobs that used to require five people. That one person may depend on a dozen online tools to do that job, and sure enough, if just one of those tools goes down, it creates a backlog. The employee will have to adjust, there may be some things she or he can't accomplish until the tool gets back up, and tasks take longer. This means the next person waiting on those results has to wait longer. The cycle repeats and a form of virtual gridlock ensues down the line. If the fix requires getting some portion of a network up and running properly, that network slows down or stops until the problem is pinpointed and solved.
Culprit No. 4: the useless knowledge ase(KB) and other ill-conceived, half-assed attempts to create useful resources. I mean, what the fuck, people? There are companies all over Silicon Valley and elsewhere whose entire raison d'être is creating searchable databases. Yet the business world is littered with half-baked, half-finished KBs that are utterly useless, outside of management pointing to the KB and pretending it's functional. This means employees can't lay their hands on information they need when they need it, which means they have to stop whatever else they're doing and go on a wild online goose chase and/or beg other overworked colleagues for help. Gridlock ensues again.
Lastly, no Sunday Kos essay would be complete without an obligatory swipe at corporate America. This means those shot-calling, overpaid, bean-counting, steely-eyed granny starvers who are convinced to a metaphysical certainty that they and they alone are paid zillions because of their sheer brilliance and admirable work ethic. They are equally convinced their companies can do without necessary infrastructure and get by on the cheap because "our people are such damn special snowflakes and we negotiate such great hardware deals."
But the commercial category that doesn't get waylaid often enough is smaller corporations and sole proprietorships. Take, for example, the tiny to medium-sized boutique operation with a few to several dozen employees that develops many of the niche applications and network tools your company eventually comes to rely on every day. Some of them do pretty damn well considering their lean structures. But as a group, these businesses tend to be understaffed and can get in over their heads really quick. Yet larger corporations continue to choose them for various reasons. Chief among them: cost savings.
Network expertise is not free and keeping up is a challenge even for the S&P 500. Smaller companies don't have the revenue stream to support the best technology, and they're often founded and run by people and close cronies who have run headlong into a brick wall crafted by Mr. Peter Principle. One network engineer friend described it thus: a company run by starry-eyed Steve Jobs wannabes who depend on and push abysmally under-powered software solutions, akin to a fleet of four-cylinder economy cars being used to haul 18-wheeler loads up mountainsides. All that so Bill Lumbergh's stock will go up a quarter point.
As a relative newcomer to the tech/hardware/software universe, I can't say if any of the ideas above hold water. There are probably bits and pieces that ring true while others ring hollow. But whatever it is that's bogging our networks down and causing crashes, we'd better get it together. Because I can tell you, functionally if albeit anecdotally, our networks just plain suck and they're getting worse all the time.
And some day soon a network issue could reach beyond being able to clock in, or pull up an invoice. It could mean driverless cars smashing into day care centers, or pilotless aircraft landing upside down on the factory that made them.