The beginning of a new year normally is a time to reflect; it's all the more so when the network industry is facing (along with the rest of the economy) a major financial crisis. We are of a generation that has taken network growth for granted, that has seen the Internet reshape culture and that has come to believe that "more bits" paves the road to the future. It just might be that this comfortable view is the greatest threat we face, because networking is going to change one way or the other.
If you look at how enterprises have spent on networks during the last couple of decades relative to their total spending on IT, you see a clear behavior shift. For most of the 1990s, networking got a larger-than-average share of investment. There was a sharp reversal in 2000, however, and since then, computer systems and software have gotten the lion's share. In the last four years, networking has steadily lost influence as a driving factor in enterprise productivity planning, my surveys of enterprises have shown.
The long timelines of this data show we're not dealing here with a cyclical modernization process. The fact is that software, servers and other computing tools are getting more attention, more budget money, more respect than networking. The fact that the point where the shift occurs corresponds to the last major economic downturn raises some legitimate questions about whether networking might not take a further hit in the current slump, as well as questions of what might be done to prevent that.
To start off, I do not believe that because networking collapsed in the post-bubble period, it has to collapse again. The major IT spending cycles of the '60s and the '80s were driven by changes in computing, and it was the latter (a distributed-computing cycle driven by the growth of PCs in business applications) that created the network boom of the '90s. It makes sense to say that distributed computing means distributing computers, which then means networking to keep them connected. Networking was catching up, and it's not surprising that IT then took the lead again.
No, but it was disappointing for sure. The question we might ask is why networking couldn't capitalize on the attention it received. The answer, I think, lies in the stuff that binds networks to applications. The pivotal point in that critical issue came in the early 1990s, when IBM's Systems Network Architecture was supplanted by TCP/IP. SNA network equipment was just too expensive, and enterprises went to the lower cost of TCP/IP instead. The critical thing was that SNA was an application architecture as well as a network architecture, and TCP/IP vendors didn't present application tools. I remember early router applications transporting SNA traffic because that's what the applications generated. Even today, the trend to service-oriented architecture (SOA) is a trend not specifically to TCP/IP but to different application-layer tools -- tools, I might add, that IBM and Microsoft and Oracle and SAP and the like are the ones providing.