TABLE OF CONTENTS The process known as "routing" in GMPLS is really not routing at all, but the distribution of information that will be used as the basis of the path computation that determines how LSPs will be placed within the network. This chapter introduces the concepts of GMPLS routing, focusing on the protocols used and the information distributed. Advanced discussion of what traffic engineering means in a GMPLS network and how paths are computed is deferred to Chapters 8 and 9.
GMPLS routing information distribution is based on extensions to IP routing protocols. Note that traffic engineering information distribution is currently limited to within an IP routing area because there are two IP routing protocols that interoperate in a scalable way within an area (OSPF and IS-IS), both of these protocols were extended by the IETF. This chapter introduces the extensions to the protocols in an abstract way before describing how the individual protocols were extended.
In an IP network, routing is the process of determining the next hop for an IP packet on the shortest path toward its destination. This decision is made by each router in turn as the packet progresses through the network and is based on information in the routing table that is either manually configured or built by routing protocols.
The chief routing protocols used within an area (OSPF and IS-IS) are link state protocols. Each router is responsible for distributing information about itself and its interfaces (that...
