What Is EIGRP Enhanced Interior Gateway Routing Protocol?

Enhanced Interior Gateway Routing Protocol (EIGRP)

Enhanced Interior Gateway Routing Protocol, or EIGRP, is a Cisco-developed hybrid routing protocol that uses the Diffusing Update Algorithm (DUAL) to calculate paths quickly and without loops. It includes elements of both distance-vector and link-state protocols. Both equal-cost and unequal-cost load balancing are supported, and the optimal path is determined by a composite metric based on bandwidth, delay, reliability, and load respectively. Fast network convergence for better performance is provided by EIGRP, which is effective at only sending updates when the network changes.

Cisco still maintains control over several complex functionalities, but in 2013, it released a limited version of EIGRP an open standard, enabling other vendors to adopt some of its features. EIGRP employs protocol number 88 and runs on the OSI model’s Network Layer Protocol.

Function and Goals

By determining the optimal route for packet delivery between any two layer 3 devices (such as routers or layer 3 switches), it automates routing decisions and configurations on computer networks. It is an intra-AS routing protocol since it operates inside a single autonomous system (AS). In 1993, EIGRP took the place of the Interior Gateway Routing Protocol (IGRP), mainly due to IGRP’s inability to handle classless IPv4 addresses.

Important Features and Ideas

Diffusing Update Algorithm (DUAL)

Based on research from SRI International, EIGRP employs the Diffusing Update Algorithm (DUAL) to increase protocol performance, avoid calculation errors, and guarantee loop-free routing throughout the whole route computation process. By facilitating speedy synchronization across routers impacted by a topology change, DUAL makes rapid convergence possible. DUAL looks for a viable successor (backup route) in the topology table and utilizes it right away if a best path (successor) fails; if none is found, EIGRP asks its neighbours for a different route.

Reliable Transport Protocol (RTP)

To assure the ordered and guaranteed delivery of EIGRP packets to all neighbours, EIGRP uses Cisco’s proprietary RTP. Both unicast and multicast packets are supported by RTP (to 224.0.0.10 for IPv4 and FF02::A for IPv6). Reliability is only offered when required for efficiency; for example, multicast hello packets do not need acknowledgements, but updates, enquiries, and replies do. Unicast can be delivered up to 16 times before a neighbour is deemed dead if a multicast fails.

Also Read About What is IP Routing, How IP Routing Works, and Types

Protocol-Dependent Modules (PDMs)

EIGRP uses PDMs to implement distinct routing procedures for several network layer protocols, including IP, IPv6, IPX, and AppleTalk. Protocol-specific requirements, such as encapsulating EIGRP packets in IP, are handled by these modules.

Neighbour Discovery/Recovery

By delivering brief Hello packets on a regular basis, routers dynamically discover information about other routers that are directly connected to them. These are “keep-alive” messages that are used to find and recover neighbours. A router considers a neighbour to be alive and operational as long as hello packets are received; only then can routing information be shared.

• Hi, Timer: EIGRP automatically delivers hello messages every 60 seconds on low-bandwidth multipoint lines and every 5 seconds on high-bandwidth links.
• Dead Timer (Hold Time): If no hello packets are received within this time, a neighbour will be deemed dead. By default, it is 180 seconds for low-bandwidth links and 15 seconds (three times the greeting interval) for high-bandwidth links. Hello packets advertise the hold time.
• Neighborship Criteria: Subnet masks, Autonomous System (AS) numbers, and K-values (metric weights) must all match for EIGRP neighborship to form. If used, MD5 and SHA-2 authentication must also match.

EIGRP Tables

Three separate tables are kept up to date by EIGRP to hold data:

• The IP addresses and interfaces of routers that are physically connected to one another are included in the neighbour table. Additionally, it contains data for dependable transportation, including Round Trip Timers (SRTT, RTO), retransmission lists, and sequence numbers.
• All known routes, including their metrics, viable successors, and successors, that have been discovered through neighbour routing tables are stored in the topology table. Routes are labelled as “active” (currently figuring out the optimum path) or “passive” (path determined). Direct traffic forwarding does not use the topology table.
• The best routes, referred to as successors, that DUAL chooses from the topology table and uses to forward traffic are listed in the routing table.

Also Read About UHF Connectors, UMCC Connector & K Connectors in Network

EIGRP Metrics

To find the optimal route to a destination, EIGRP employs a composite measure. Although up to five variables can be employed in the composite metric calculation, only bandwidth (K1) and latency (K3) are used by default. The remaining variables are Minimum Transmission Unit (MTU) (K5), load (K2), and dependability (K4).

• Bandwidth: The path’s minimum bandwidth (measured in kilobits per second) from the source to the destination.
• Delay: The total delay of all the path’s links, measured in tens of microseconds.
• Load: A value between 1 and 255, where 255 indicates saturation.
• Reliability: A value between 1 and 255, where 255 indicates the highest level of reliability.
• MTU: Maximum Transmission Unit on the Minimum Path. The metric calculation never makes use of this.

Since K1=1, K2=0, K3=1, K4=0, and K5=0 are the default K-values, the formula becomes (Bandwidth + Delay) * 256. To create an adjacency, these K-values must coincide on every router in an EIGRP system.

Feasible Successors and Loop Prevention

A successor is the nearby router used for packet forwarding that provides the cheapest, loop-free route to a destination. A second next-hop router that offers a loop-free route to the same location-possibly over a greater distance-is a viable replacement. In EIGRP networks, the Feasibility Condition is a sufficient condition for loop freedom: a neighbour is assured to be on a loop-free route if they advertise a distance to a destination that is strictly less than the feasible distance (current best metric) of the local router. Fast convergence results from EIGRP’s ability to quickly transition to a viable successor in the event that one becomes unavailable. Other loop prevention techniques like split horizon and poison reverse are also employed by EIGRP. Split horizon stops promoting a way to leave the interface where it was learnt. Using the same interface, Poison Reverse promotes a learnt path as unreachable. These rules are combined by EIGRP, which applies them to queries, topology changes, and startup.

Packet Types

EIGRP communicates with neighbours using five different message types:

• Greetings messages: used for unicast acknowledgements, neighbour detection, and recovery.
• Update messages: Indicate whether a destination is reachable. During topological changes (partial updates), they either solely include new routes or all of the best routes. Updates are consistently given.
• Query messages: Multicast when a router has to locate a different path or when a device is deemed dead and has no routes in its topology table. Reliable queries are sent.
• Messages that acknowledge query messages and provide the path to the network being requested are known as reply messages. Reliable and unicast responses are sent.
• It updates, questions, and responses are acknowledged by acknowledgement messages. Acks are unicast hello packets that are empty of data.

States

There are two possible states for a destination in the topology table:

• When the router has determined the destination’s successor or successors, the state is said to be passive.
• Active state: When no viable successors are found and the current successor no longer meets the feasibility requirement, the router is going through a route recomputation. The router is unable to switch its next-hop neighbour for packet forwarding during this period. Neighbour session resets result from the route becoming Stuck in Active (SIA) when a query remains unanswered for an extended period of time.

Also Read About IGP Interior Gateway Protocol Types, Purpose, And Features

Load Balancing

Both equal-cost and unequal-cost load balancing across several routes to a destination are supported by EIGRP.

• By default, it can add up to four equal-cost routes to the routing table (or up to 16 in more recent Cisco IOS versions).
• The variance command configures unequal-cost load balancing. Traffic is placed on any link whose metric is less than the metric of the optimal path multiplied by the variance. The variance is a multiplier (default 1, max 128) that decides which paths to include.

Route Summarization

Both manual and automatic route summarization are supported by EIGRP.

• By default, automatic summarization is enabled, aggregating routes at class borders and preventing loops by installing a summary route with a Null0 interface as the next hop. Without auto-summary, it can be turned off.
• The ip summary-address eigrp command is used to configure manual summarization at the interface level, enabling summarization on almost any bit boundary.

Administrative Distance (AD) The default administrative distances in EIGRP routing information are:

• 90 internal routes.
• External Routes: 170 (for routes redistributed from other protocols or learnt from an EIGRP-enabled router outside the autonomous system).
• Routes in summary: 5.

Hop Count

Although it is not utilized in the computation of metrics, EIGRP keeps track of the number of hops on each route. It is possible to change the default maximum hop count of 100 to 255. Routes that go beyond this point are marked as inaccessible.

EIGRP for IPv6 (EIGRPv6)

EIGRPv6 is the new version of EIGRP that supports IPv6.

• Many of the characteristics of EIGRP for IPv4 are still present, including DUAL, RTP, and quick, loop-free convergence.
• FF02::A is the multicast address for EIGRPv6 Hello and update packets.
• The fact that EIGRPv6 does not use network statements is a significant configuration difference. Rather, interface-specific commands (such as ipv6 eigrp [ASN]) are used to include interfaces in the routing process, and the routing process itself requires a no shutdown command in router configuration mode.

Also Read About Trunking Protocol: A Simple Solution For Complex Network

Configuration Basics (IPv4 example)

To make a Cisco IOS router EIGRP-capable:

Router# configure terminal

Router(config)# router eigrp [autonomous-system-number]

Router(config-router)# network [network-number] [wildcard-mask]

Router(config-router)# no auto-summary

Every router in the same EIGRP domain needs to have the same Autonomous System (AS) number.

Advantages of EIGRP

Rapid convergence was made possible by DUAL and workable successors.

• Decreased Bandwidth Usage: Sends triggered, partial updates rather than whole routing tables on a regular basis.
• Scalability: Effectively determines the optimal paths in intricate and sizable networks.
• Multiprotocol Support: enables AppleTalk, IP, IPv6, and IPX through PDMs.
• Cost load balancing: A special feature that permits traffic distribution across paths with varying costs is unequal-cost load balancing.
• Loop-free topology: The feasibility condition and DUAL ensure a loop-free topology.
• Compatibility: IGRP is backwards compatible.
• Flexible Routing Decisions: Enables link characteristics to be adjusted using metrics.

Limitations of EIGRP

• Proprietary Nature: Originally Cisco-proprietary, it remains strongly Cisco-centric even though it is now an open standard. The RFC’s omission of essential information can make interoperability with other vendors difficult.

• Complexity: Careful configuration may be necessary for advanced functionality.

• Resource-intensive: May use more memory and CPU than less complex protocols, particularly while recalculating routes.

All things considered, EIGRP is a reliable and effective routing protocol that works especially well for Cisco-based enterprise networks that need quick convergence, economical bandwidth use, and adaptable traffic control.

Also Read About What Is Dynamic Trunking Protocol CISCO And Characteristics