A routing protocol is an algorithm that gives a Router the ability to learn (dynamically) about specific existing networks around it (or even the Internet). Practically, these routing protocols help routers learn dynamically about each other without the need for an administrator to intervene.
This helps us primarily because it does not require an administrator to set up each network individually.
When talking about dynamic routing protocols, we generally refer to 2 classes:
1.Distance Vector (RIP, EIGRP)
2.Link State (OSPF, IS-IS).
Distance Vector are those routing protocols that are based on the “Routing by rumor” principle (that is, they rely on the information received from my neighbor, and do not have an overview of the entire network).
Link State are routing protocols that have an overview (an active map) of the entire network. They know “everything that moves.” They have information about each link (its current status: up/down, the speed of that link – 100Mbps, 1Gbps, and so on).
Here are some examples of routing protocols in Routers:
OSPF
OSPF or Open Shortest Path First, is a routing protocol that is part of the Link-State category, unlike RIP that is a Distance Vector. Specifically, OSPF is the most widely used routing protocol. Appears in the vast majority of networks, from medium to large. The main reason being that it is vendor-independent and that can be used by anyone.
The purpose of this routing protocol is to find the fastest route to your destination. OSPF calculates this path based on a cost defined as follows:
Cost = 1000/BW, where BW represents network speed – 10, 100, 1000 Mbps.
Thus each network has a specific cost (100, 10, 1 depending on the speed), and the total cost from a source network A to a destination network B will be the sum of costs (accumulated costs).
EIGRP
2.EIGRP or Enhanced Internal Gateway Routing Protocol is a routing protocol developed by Cisco for Cisco devices.
EIGRP is a routing protocol created by Cisco to become the fastest and easiest direct competitor to OSPF. It is part of a particular category called Advanced Distance Vector because it borrows features from both routing protocols like RIP (Network Learning Module From Neighbors, respectively OSPF Update Packets).
BGP
BGP (Border Gateway Protocol) is the routing protocol used in the Internet core. It maintains a table of IP networks (or “prefixes”) that show the path used to reach a specific network through the various autonomous systems (AS). BGP is therefore considered a vector-path routing protocol (as opposed to vector-distance protocols that do not map the distance all the way). BGP does not use the same metrics as the routing protocols used within ASs but takes decisions based on the route and routing policies of the stand-alone system it is part of.
The protocol was created to replace another routing protocol EGP and to allow the decentralization of routing on the Internet, making its old NSFNet network core obsolete. Since 1994, the fourth version of the protocol is used on the Internet, with all previous versions being considered obsolete. The most significant progress of version 4 was the added support for CIDR and the use of route aggregation to reduce the size of routing tables.
Choosing the route
The routing algorithm extracts the destination IP address from the IP packet, then checks whether that address corresponds to any of its interface addresses. If not, it sequentially navigates the routing table by comparing the result of the logical operation performed between the destination IP address and the network mask extracted from the routing table entry. If the result of the logical operation corresponds to the network address in the routing table, the IP packet is passed to the next-hop IP. If none of the routing table networks match the destination address, the package is ignored.