The comprehensive, hands-on guide for resolving IP routing problems:
The primary objective of this site is to provide elaborate guidance for troubleshooting Internet
Protocol (IP) routing problems on Cisco routers. In this regard, the subsequent text covers wellknown
routing protocols such as the following:
- Open Shortest Path First Protocol (OSPF)
- Integrated Intermediate System-to-Intermediate System Protocol (IS-IS)
- Border Gateway Protocol (BGP)
- Protocol Independent Multicast (PIM) for multicast routing
The Transmission Control Protocol/Internet Protocol (TCP/IP) suite of protocols is the underlying
technology for information exchange on the Internet. TCP/IP uses a layering approach for computer
communications similar to the Open System Interconnection (OSI) reference model, but with fewer
than seven layers. Figure 1-1 shows the OSI reference model and the TCP/IP stack side by side.
Related layers between the two stacks are indicated in the figure.
Figure 1-1. OSI Reference Model and TCP/IP Stack

IP operates at the Internet layer of the TCP/IP suite, which corresponds to the network layer of the
OSI reference model. IP provides connectionless data-delivery services, which involve transmission of
information from one part of a network to another in units of data known as packets or datagrams.
The essence of the datagram delivery service model is that a permanent pre-established end-to-end
path is not required for data transfer between two points in a network. In a packet-based network,
each router in the transmission path makes independent local decisions regarding the optimal
forwarding path toward the destination for any transit packet. The decision-making is based on forwarding intelligence gathered either dynamically by means of a routing protocol or manually
programmed static routes.
Addressing is an important aspect of the data-forwarding process. For any directed com-munication,
there is a source and a destination. Addressing allows the target destination to be specified by the
source and allows the destination node to also identify the source. Addressing is even more important
in the datagram delivery mode of operation because, as in IP forwarding, the data path for any
transmission is not nailed through the intermediate nodes between the source and destination.
As mentioned previously, within the IP datagram services infrastructure, information that is to be
transmitted from one device to another first is broken down into packets. Each packet has an IP
header, a transport layer (TCP or UDP) header, and a payload, which is a piece of the original
information. Each IP packet is self-contained and independently is forwarded to the destination
through the chain of intermediate devices that might be along the path of transmission.
The routers in the network depend on a routing protocol or static configuration to forward the
datagrams in a stream to their intended destination. For any destination address, each node in the
data path worries about only the outgoing interface or link along a locally determined optimal path to
the destination (or as specified by a special forwarding policy). The IP for-warding process frequently
is described as a hop-by-hop destination-based forwarding mechanism. This means that routers at
each hop along the data path normally forward packets based on the destination address. However,
modern routers also can use policy-based criteria, such as the source address in a packet to direct
the forwarding.
At the destination, packets belonging to the same stream are reassembled into the original
information. IP addressing is discussed in the next section,
IP Addressing Concepts.
This process of forwarding a packet from one node to the other in a connectionless network based on
the Layer 3 address (IP address, in this case) also is referred to as routing. Routers are specialized
network devices with acquired routing intelligence.
So how do routers really decide where and how to forward packets traversing the inter-network?
Well, this is done in a couple of ways. As alluded to previously, routers can be manually
preprogrammed with predetermined path information known as static routes, or they can run
applications that facilitate the learning and sharing of routing information automatically. Obtaining
and propagating routing information by the latter method is re-ferred to as dynamic routing.