The recent three-day unprecedented outage of Skype services highlights some interesting characteristics of contemporary networks. Does it matter what really caused it? Probably not. Because the real issue is that we don’t fully understand modern digital networks. Whether or not you accept the Skype line that it was all triggered by “a massive restart of users’ computers across the globe within a very short timescale”. Or whether you prefer to believe the inevitable accusations of rivals that it was all down to fundamental flaws in their systems. The problem is that large digital networks are a law unto themselves. They are often unpredictable and they frequently exhibit behaviour that appears to be self-generated.
Hub-and-spoke networks are particularly hard to fathom, because they possess entirely different topological (and other) characteristics from traditional point-to-point organic networks. Important characteristics such as performance, security and failure rates can be very, very different in these so-called, scale-free networks. You need to be an expert in complexity theory to gain any insight into what’s really going on.
Customer behaviour can also generate strange effects. Users might, for example, generate a huge increase in transactions when response times are slow by constantly pressing the send key. Collaborative, simultaneous network effects are also possible. Put all this together and we can expect some interesting times as organisations move towards increasing dependency on large-scale, hub-and-spoke digital networks.
In my view we need a lot more research and much better education to understand the real consequences of managing modern digital networks. I’ve occasionally pointed out some of these problems to business managers responsible for implementing new hub-and-spoke networks. Their reaction? Rather like a frightened rabbit in headlights.