The routing protocol is the part of a network stack that enables data to get to its ultimate destination in the most efficient way. They function by checking for optimal routes, building routing tables, and deciding on routes for your traffic. Without a routing protocol, packets wouldn't get directed to where they need to go.
There are various routing protocols, and they all work slightly differently but basically do the same thing; they provide addressing information to allow data to be forwarded from one system or network to another as seamlessly as possible.
In this article, we will look at the most prominent routing protocols and explain why they can each be categorized as either:
- Link state or distance vector
- IGP or EGP
- Classless or Classful
Types of routing protocol
There are many routing protocols that perform packet switching functions during networking. We can place the vast majority of those protocols into two main classes:
- Link state protocols
- Distance vector protocols
Link state protocols
Link state protocols function by communicating from router to router in order to decide on the best possible route for packets.
It is a fast protocol because it doesn't rely on routers to advertise a routing table. Instead, link-state routers multi-cast Link-State Advertisements to all neighboring routers only when a change occurs, which saves on bandwidth.
Link state protocols leverage statistical information such as speed, cost, and congestion to determine the best routing path. In addition, they use variable-length subnet masks, which makes addressing more efficient because of scalability.
The most common types of link-state protocols are:
- IS-IS (Intermediate System-Intermediate System)
- OSPF (Open Shortest Path First)
- NLSP (NetWare Link Services Protocol)
One of the big advantages of the link-state protocol is that it does not suffer from routing loops, an error that affects some routing protocols and causes a data packet to be continually routed through the same routers again and again.
Distance vector protocols
Distance vector protocols function by constantly advertising the entire routing table during updates. They generally send out updates once every 30 to 90 seconds.
Distance vector protocols are relatively basic, and primarily leverage distance, or hop count (the number of intermediate network devices through which data must pass) to determine the best route for forwarding packets.
Distance vector protocols are vulnerable to routing loops
Unlike the link-state protocol, which only sends out a multi-cast update when routers are triggered to do so because of a change, distance vector protocols broadcast updates periodically and regardless of change – which unnecessarily consumes bandwidth.
The most common types of distance vector protocols are:
- RIP (Routing Information Protocol)
- IGRP (Interior Gateway Routing Protocol)
- EIGRP (Enhanced Interior Gateway Routing Protocol)
Link state protocols
Here's a breakdown of the most common link-state protocols.
Intermediate System-Intermediate System
IS-IS is an interior gateway protocol (IGP) that uses link-state information to decide how to route packets. It leverages the shortest-path-first (SPF) algorithm (a modified version of the Dijkstra algorithm) to determine optimal routes using the link-state database.
The IS-IS protocol is most commonly used as an IP routing protocol on the internet. It uses ISO network addresses (NSAP and NET), which means that the configuration of both IPv4 and IPv6 is the same.
An IS-IS network is an Autonomous System (AS) made up of end systems (user devices that send and receive packets), and intermediate systems (routers). In an IS-IS network, routers are organized into groups known as areas, and multiple areas are grouped together into domains.
They break routing down into two hierarchical classes; Level 1 and level 2. Level 1 routers share routing information within an area (intra-area), and Level 2 routers share information about IP addresses to connect those areas together (inter-area).
Open Shortest Path First
OSPF is a link-state IGP that was developed specifically for IP networks. It uses the shortest-path-first (SPF) algorithm to choose the best route for packets.
OSPF routers all store a copy of the link-state database (network topology maps of the distances to other routers on the surrounding network). They analyze that map o determine the shortest path through the network.
The primary difference between OSPF and routing protocols that came before it (such as IS-IS and the Routing Information Protocol) is that it calculates the shortest path through a network while also considering the link cost, bandwidth, congestion, and load.
Due to its ability to reliably calculate routes through large and complex LANs, and its ability to recalculate compromised routes, OSPF has been widely adopted by many organizations.
NetWare Link Services Protocol
NLSP is a routing protocol for Inter-network Packet Exchange. It is based on the IS-IS protocol developed by the ISO. NLSP was designed to overcome the shortcomings of the RIP and SAP protocols that it replaces.
It is a link-state protocol designed specifically for NetWare networks, that permits for the exchange of routing and service information without high broadcast overheads.
One of the primary benefits of NLSP over its predecessors is that it doesn't constantly retransmit its information every few minutes. Instead, it transmits when there is a change in a route, or once every two hours, which makes it better suited for use over a Wide Area Network.
Distance vector protocols
Here's what you need to know about the most common distance-vector protocols.
Routing Information Protocol
RIP, RIPv2, and RIPng are distance vector routing protocols that use hop count as the routing metric to decide the best networking path for packets. RIP uses port 520, and it works on the application layer of the OSI model.
In a RIP network, the path with the lowest number of hops is always considered optimal. No other metrics, such as link cost and bandwidth, are considered. The maximum number of hops permitted is 15, which rules RIP out for use over a Wide Area Network.
The primary distinction between RIP and its two predecessors (RIPv2 and RIPng) is that RIP broadcasts messages to all connected hosts. The two latter versions use multicasting to send the information only to the intended recipient, which creates less traffic and makes the system faster.
The drawback of the RIP protocol is that it broadcasts routing information every 30-seconds to all connected hosts; even if there are no changes to report. Those updates include full routing tables. As a result, the protocol results in increased network traffic.
In addition, RIP cannot support variable-length subnet masking (to divide an IP address space into a hierarchy of subnets).
Interior Gateway Routing Protocol
IGRP is a protocol that was developed by Cisco in 1985. It is a distance-vector protocol developed to build on the foundations provided by the RIP protocol.
It improves on its predecessor by permitting more than 15 hops, which makes it suitable for larger connected networks. In fact, it increases the maximum number of hops to 255, which makes it much more versatile.
By default, IGRP uses bandwidth and delay metrics to calculate the best possible route for packets. However, it can be set up to consider other metrics too, such as reliability and load.
IGRP is resistant to routing loops, and updates automatically when any route changes take place within the network
The biggest drawback is that IGRP is a proprietary protocol developed by Cisco that only works with Cisco routers. Another disadvantage is that (like RIP) it broadcasts the entire routing table every 90 seconds, which consumes a lot of network bandwidth.
Enhanced Interior Gateway Routing Protocol
EIGRP is a distance-vector protocol that first became available in 1992. It builds on its predecessor IGRP, and, like it, it is a Cisco developed proprietary protocol. EIGPR is leveraged by IP networks, AppleTalk, and NetWare networks.
On an EIGPR network, routers get information from their neighbors to access, share, and record routing table information. In practice, this results in each router query neighboring routers for route information, and if a change has occurred, they pass the information along so that the neighboring router can update its table.
One of the advantages of EIGPR is that despite being a distance-vector protocol, it is actually very efficient. As a result, it causes low usage of network resources. This is because only routing changes, rather than the while routing table, are broadcast.
Another benefit of this enhanced protocol is that it uses the Diffused Update Algorithm (DUAL) to calculate the shortest path to a destination, and to allow for recalculation that prevents the possibility for routing loops and permits fast convergence.
IGPs and EGPs
In addition to being categorized as either distance vector or link-state protocols (depending on how they function), routing protocols also fall into two other categories:
- IGPs (Interior Gateway Protocols)
- EGPs (Exterior Gateway Protocols )
The important distinction between these two protocols is that an IGP is used to transmit data between routers that exist on the same Autonomous System (AS). (Where an AS is defined as a network or collection of networks that are all controlled by one single entity, organization, or company.) For example, a company's network is a separate AS to an Internet Service Provider's network.
Of the routing protocols we have covered in this guide, the following are IGPs:
- OSPF (Open Shortest Path First)
- RIP (Routing Information Protocol)
- IS-IS (Intermediate System to Intermediate System)
- EIGRP (Enhanced Interior Gateway Routing Protocol)
An EGP, by contrast, is used to transmit data between independent Autonomous Systems. They are more complex in nature, and Border Gateway Protocol (BGP) is the most commonly used EGP. That said, there are a few EGPs available to network administrators:
- BGP (Border Gateway Protocol)
- EGP (Exterior Gateway Protocol)
- IDRP (The ISO's InterDomain Routing Protocol)
Border Gateway Protocol
BGP first became available in 1995 when it was released to replace EGP. It is an innovative protocol that leverages a decentralized approach to routing. BGP picks the best route for packets by leveraging the Best Path Selection Algorithm.
The benefit of BGP is that network administrators can easily elect to bypass automated routing decisions made by the protocol in accordance with their specific needs and desires.
Thus, BGP is an advanced protocol that can decide routes based on additional advanced factors (such as weight, local preference, locally generated, AS Path length, origin type, multi-exit discriminator, eBGP over iBGP, IGP metric, router ID, cluster list and neighbor IP address). And as a result, an EGP permits for global network connectivity based around administrator-defined policies.
Another major benefit of BGP is that it is provides high stability, and can be setup with authentication to ensure that only verified hosts may communicate with BGP routers.
Classless VS Classful protocols
Besides being classified as either EDP or IGP protocols and link-state or distance-vector protocols, routing protocols can also be defined as either Classless or Classful. This categorization refers to the method by which routing protocols perform routing updates.
Classless
- RIPv2
- EIGRP
- OSPF
- IS-IS
Classless protocols send IP subnet mask information during routing updates and support Variable Length Subnet Masks and improved route summarization. It also allows for efficient address space allocation and eliminates class imbalances.
Classful
- RIP
- IGRP
Classless protocols do not include subnet mask information during routing updates. A drawback of Classful protocols is that they broadcast periodic updates to all connected interfaces which causes congestion.