Dynamic routing protocols

Static routing is used :
1. if network is small
2. if there is a single connection point to other networks
3. if there are no redundant routes
Dynamic routing is used :
if any one of above conditions fail.
Under dynamic routing, routers communicate to each other, informing each other about the networks, they are currently connected to. Routers communicate with other routers within an autonomous system via. interior gateway protocols (IGPs). Router from one autonomous system communicate with router in other autonomous systems via. exterior gateway protocols (EGPs).

Interior Gateway Protocol

Routing Information Protocol (RIP)
It is a distance vector protocol. i.e. messages sent by RIP contain 'a vector of distances' or 'hop counts' . Each router updates its routing table after receiving hop counts from its neighbour router.

Open Shortest Path First (OSPF)
It is a newer alternative to RIP. It is a link state protocol. It does not exchange vector distances with its neighbours. Each router tests status of its link with neighbour routers and prepares its routing table.

Exterior Gateway Protocol

Border Gateway Protocol (BGP)
Works with routers belonging to two different autonomous systems. It exchanges network reachability information. That is, the full path from one network to other.

IP routing

A datagram is generated by either a localhost or any other host. When this datagram seems to generate from any host other than localhost, then the host may work as a router. These hosts contain routing tables which help route datagrams. When any datagram is received by a host and not belong to it, gets dropped (silently discarded). This routing of datagram done by host as router is IP routing. Lets see how routing is done.
IP routing:
A datagram is either generated or forwarded by a system. A search for routing table is done for this generated datagram. This table is updated by either a routing daemon or ICMP redirects. Static routes can be entered by using the route command. ICMP redirects updates routing table dynamically.

Traceroute Program

The Traceroute program was written by Van Jacobson. When an IP datagram moves from one host to another, Traceroute program lets us see the route that IP datagram follows. It is not always true that two consecutive IP datagrams follow the same route when they move from the same source to the same destination but most of the time they follow the same route. We can be aware of this with Traceroute program.
Operation of Traceroute program:
This program send UDP datagrams starting with a TTL (time to live) of 1. This TTL with value 1 finds a router and gets destroyed returning 'ICMP time exceeded' error. Now another datagram with TTL 2 is sent. It lives till it encounters 2 routers returning same error. Now another TTL with value 3 is sent. Thus each time the TTL is increased by 1, to locate each router in the path. An 'ICMP time exceeded' is returned by each router when it discards the UDP datagram, and an 'ICMP port unreachable' is generated by the final destination.

Ping Program

The Ping program was written by Mike Muss. Ping program tests the connectivity between two systems running TCP/ IP. With it one host may test whether another host is reachable. It uses  the ICMP echo request and echo reply messages. The program sends an ICMP echo request message to a host, expecting an ICMP echo reply to be returned. Thus, connectivity is confirmed.

Uses of ping:
1. To measure round trip time to the host.
2. To examine and use the IP record route option.
3. To examine the IP timestamp option.

ICMP

ICMP (Internet Control Message Protocol) is a part of IP layer. ICMP messages are transmitted within IP datagrams. An IP datagram consists of an ICMP message with IP header.
ICMP communicates either error messages or query messages. To communicate a message it has a type field and a code field. A type field consists of 15 different values viz. 0, 3, 4, 5, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, to identify the type of ICMP message and a code field to specify the condition following the type. e.g. An echo request (ping request) has a type field value 8 and a code field value 0.

Reverse Address Resolution protocol

Reverse address resolution protocol does physical address to logical address mapping. A system with a local disk obtains its IP address from a configuration file when bootstrapped. But RARP is used by diskless systems such as an X terminal or a diskless workstation, to obtain their IP address when bootstrapped.
ARP and RARP both broadcast the request, and receive unicast reply. RARP request is broadcast with sender's hardware address, asking for sender's IP address.

Address Resolution Protocol

Suppose we send an Ethernet frame from our pc on a LAN to another pc. A data link such as an Ethernet or a token ring has its own addressing scheme which is often  48-bit addresses. Information is sent using IP addresses which is 32-bit address. Address Resolution Protocol provides a mapping between two different forms of addresses: 32-bit IP addresses and whatever type of address the data link uses (here 48-bit ethernet address). This logical address to physical address mapping is done by ARP.
ARP does dynamic mapping of addresses. That is, it is neither a concern of the application user or the system administrator to configure anything when this mapping takes place. All work is done automatically.