So far in this book, we have made a distinction between real-time and nonreal-time traffic as the major differentiating factor regarding the tolerance toward delay, jitter, and packet loss. We have also kept the traffic flows simple, just a series of Ethernet, IP, or MPLS packets crossing one or more devices.
This is clearly an over simplification, so in this chapter we take one step further by examining some of the more special scenarios. The environment where a QOS deployment is performed, either in terms of the type of traffic and/or type of network present, always needs to be taken into consideration. However, some environments are more special than others due to their uniqueness or just because they are a novelty; as such the authors of this book have selected the following scenarios as the most “special ones” in their opinion, and each one will be detailed throughout this chapter:
- Layer 4 transport protocols—The User Datagram Protocol (UDP) and Transmission Control Protocol (TCP). Understanding how traffic is transported is crucial, for example, if the transport layer can retransmit traffic, is packet loss really an issue? Or how does the transport layer react when facing packet loss? How can it be optimal?
- Data Center—An area where the developments in recent years mandate a special attention. We will analyze storage traffic, the creation of lossless Ethernet networks, and the challenges posed by virtualization. Also there’s the current ...