The All-New Switch Book: The Complete Guide to LAN Switching Technology, Second Edition

6.4. Source-Routed Frames

Stations that implement source routing insert routing information into all frames that must traverse multiple rings. Thus, the frame format differs between bridged and non-bridged traffic.

6.4.1. Differentiating Source-Routed and Non-Source–Routed Frames

Source routing operation was fully defined after the development of the underlying Token Ring and FDDI technologies over which it operates. Thus, there was originally no accommodation made within the respective frame formats to allow a bridge to easily distinguish between source-routed and non-source–routed frames. Such differentiation is clearly needed — a key factor in the design of source routing is that bridges need only consider those frames that contain routing information; we do not want source routing bridges to have to operate in promiscuous mode.

In looking for a place to squeeze additional information into the frame, we can observe that, while a frame may be sent to either a unicast or a multicast destination, frames are always sent from an individual station. The first bit of the Destination Address indicates the unicast or multicast nature of the recipient(s). When used in its originally intended manner, the first bit of a Source Address should always be 0, indicating an individual sending station. Source routing uses this "wasted" bit to indicate the presence of source routing information, as shown in Figure 6-4. If the first bit is 0, the frame does not contain source routing information; ...