QOS series part 1: Introduction to Quality of Service
With all the recent hubub about ISP traffic management, I figured I'd start an article series on QOS: what it is, how it works, and where it is implemented in a network. After all, my day job is networking so I may as well share some knowledge.
Since the Internet first went commericial in the mid-90s, network traffic has grown significantly (that's a bit of an understatement really). These days there is talk of an impending "exaflood" of network traffic - an increase of network bandwidth of epic proportions so high that all Internet traffic becomes jammed. This is little more than a theory at this point; an earlier prediction in the 1990s of a "petaflood" never actually materialized thanks to the dizzying pace of speed increases in computing power both in generic PCs and networking equipment and the phenomenal increases in network bandwidth (in the early 1990s most of the Internet was connected by 45Mbps DS3s; nowadays there are multiple 10Gbps connections between most ISPs).
The most obvious way to manage a network is to implement quality of service. QOS is called both "managed fairness" and "managed unfairness" because it attempts to provide better service for some things while providing worse service for others.
Part 1 of my QOS series lays down the foundation of QOS. Subsequent parts will discuss technologies in more detail.
Read on dear readers and learn the fundimentals of quality of service!