Cisco EIGRP Configuration: Complete Run Book for Named EIGRP and Classic EIGRP on IOS-XE

What is EIGRP?

Enhanced Interior Gateway Routing Protocol (EIGRP) is a Cisco-proprietary advanced distance-vector routing protocol that uses the Diffusing Update Algorithm (DUAL) to guarantee loop-free path calculation. Unlike traditional distance-vector protocols, EIGRP maintains a topology table alongside the routing table, enabling fast convergence and instant failover to backup routes without recalculation.

EIGRP uses IP protocol number 88 and sends updates via multicast to 224.0.0.10. It natively supports IPv4, IPv6, and multi-protocol operation through Named Mode. This guide covers both Classic EIGRP and Named EIGRP on Cisco IOS and IOS-XE platforms including ISR, ASR, and Catalyst series.

EIGRP Terminology

• Feasible Distance (FD) — The best composite metric from the local router to the destination. The route with the lowest FD becomes the Successor.
• Reported Distance (RD) / Advertised Distance (AD) — The metric a neighbour reports for a destination from its own perspective.
• Successor — The primary next-hop with the lowest FD. Installed in the routing table.
• Feasible Successor (FS) — A loop-free backup route. Qualifies when its RD is strictly less than the current FD (Feasibility Condition). Promoted instantly to Successor without running DUAL.
• DUAL — Diffusing Update Algorithm. Guarantees loop-free topology convergence by querying neighbours only when no Feasible Successor exists.
• Active State — A route enters Active state when DUAL sends queries to neighbours to find a new path. Normal state is Passive.

EIGRP Composite Metric

EIGRP calculates a composite metric using five K-values: bandwidth (K1), load (K2), delay (K3), reliability (K4), and MTU (K5). Default K-values are K1=1, K2=0, K3=1, K4=0, K5=0 — only bandwidth and delay are used by default.

Metric = (10^7 / Min_BW_kbps + Sum_Delay_10us) x 256

Example — GigabitEthernet (BW=1000000 kbps, delay=10):
  = (10000000 / 1000000 + 10) x 256
  = (10 + 10) x 256
  = 5120

Critical: All routers in an EIGRP AS must use identical K-values. Mismatched K-values prevent neighbour adjacency formation entirely. Never change K-values without a coordinated network-wide change window.

Classic EIGRP vs Named Mode

• Classic EIGRP — Uses

  router eigrp <AS-number>

  . Separate processes for IPv4 and IPv6. Interface-specific commands (authentication, timers) are configured under the interface itself. Supported on all IOS versions.
• Named EIGRP — Uses

  router eigrp <name>

  with address-family stanzas. Supports IPv4, IPv6, and VRF within a single process. All per-interface config lives under

  af-interface

  stanzas. Supports SHA-256 authentication. Recommended for all IOS-XE 15.3+ deployments.

Classic EIGRP Configuration

The

network

statement enables EIGRP on interfaces whose IP addresses match the specified network and wildcard mask.

! rtr-infrarunbook-01
router eigrp 100
 eigrp router-id 1.1.1.1
 network 10.0.0.0 0.255.255.255
 network 192.168.10.0 0.0.0.255
 no auto-summary

! rtr-infrarunbook-02
router eigrp 100
 eigrp router-id 2.2.2.2
 network 10.0.0.0 0.255.255.255
 network 192.168.20.0 0.0.0.255
 no auto-summary

no auto-summary

is enabled by default in IOS 12.x — always disable it explicitly to prevent classful summarisation from breaking routing. In IOS 15.x and IOS-XE it is off by default. Always configure

eigrp router-id

explicitly to a stable loopback address.

Named EIGRP Configuration

Named mode centralises all EIGRP configuration — global, per-interface, and per-topology — under a single named process.

! rtr-infrarunbook-01 (Named Mode)
router eigrp INFRARUNBOOK
 !
 address-family ipv4 unicast autonomous-system 100
  !
  af-interface GigabitEthernet0/0/0
   hello-interval 5
   hold-time 15
  exit-af-interface
  !
  af-interface GigabitEthernet0/0/1
   passive-interface
  exit-af-interface
  !
  topology base
   no auto-summary
  exit-af-topology
  !
  network 10.0.0.0 0.255.255.255
  network 192.168.10.0 0.0.0.255
  eigrp router-id 1.1.1.1
 exit-address-family

Passive Interface

Suppress EIGRP hellos on interfaces that should not form neighbour relationships — such as LAN stub segments, loopbacks, and management interfaces.

! Classic mode
router eigrp 100
 passive-interface default
 no passive-interface GigabitEthernet0/0/0
 no passive-interface GigabitEthernet0/0/1

! Named mode
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  af-interface default
   passive-interface
  exit-af-interface
  af-interface GigabitEthernet0/0/0
   no passive-interface
  exit-af-interface

Using

passive-interface default

and selectively enabling active interfaces is the most secure approach. It prevents accidental neighbour formation on new interfaces and reduces EIGRP multicast traffic on access ports.

EIGRP Authentication

MD5 Authentication (Classic Mode)

! Step 1: Define key chain
key chain INFRARUNBOOK-CHAIN
 key 1
  key-string InfraRunBook@EIGRP2024
  accept-lifetime 00:00:00 Jan 1 2024 infinite
  send-lifetime   00:00:00 Jan 1 2024 infinite

! Step 2: Apply to EIGRP-facing interface
interface GigabitEthernet0/0/0
 ip authentication mode      eigrp 100 md5
 ip authentication key-chain eigrp 100 INFRARUNBOOK-CHAIN

SHA-256 Authentication (Named Mode — Recommended)

! Step 1: Define key chain with HMAC-SHA-256
key chain INFRARUNBOOK-SHA-CHAIN
 key 1
  key-string InfraRunBook@EIGRP2024
  cryptographic-algorithm hmac-sha-256

! Step 2: Apply in af-interface stanza
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  af-interface GigabitEthernet0/0/0
   authentication mode hmac-sha-256 InfraRunBook@EIGRP2024
  exit-af-interface

SHA-256 is only available in Named Mode and provides stronger authentication than MD5. Authentication failures are silent — neighbours simply do not form. Use

debug eigrp packets

to detect authentication mismatches during troubleshooting.

Verifying EIGRP Neighbours

! Active neighbours
show ip eigrp neighbors

! Detailed — uptime, SRTT, RTO, Q count, stub flags
show ip eigrp neighbors detail

! EIGRP-participating interfaces
show ip eigrp interfaces
show ip eigrp interfaces detail

rtr-infrarunbook-01# show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(100)
H   Address         Interface       Hold Uptime   SRTT   RTO  Q  Seq
                                   (sec)         (ms)       Cnt Num
0   10.0.1.2        Gi0/0/0          11 00:15:34   15   100  0  42
1   10.0.2.2        Gi0/0/1          13 00:08:17   23   138  0  31

A Q Cnt persistently above 0 indicates a congested or unresponsive link. SRTT is the smooth round-trip time used to compute the RTO (retransmit timeout).

EIGRP Topology Table

! Full topology table
show ip eigrp topology

! All paths including non-feasible successors
show ip eigrp topology all-links

! Summarised view
show ip eigrp topology summary

rtr-infrarunbook-01# show ip eigrp topology
EIGRP-IPv4 Topology Table for AS(100)/ID(1.1.1.1)
Codes: P - Passive, A - Active

P 192.168.20.0/24, 1 successors, FD is 2816
        via 10.0.1.2 (2816/2560), GigabitEthernet0/0/0   ! Successor
        via 10.0.2.2 (3072/2560), GigabitEthernet0/0/1   ! Feasible Successor (RD 2560 < FD 2816)

The route via

10.0.2.2

is a Feasible Successor because its RD (2560) is less than the Successor FD (2816). If the primary path fails, EIGRP promotes this route instantly without running DUAL queries.

Route Summarisation

Manual Summary — Classic Mode

! Advertise 10.1.0.0/22 outbound on Gi0/0/0 (covers 10.1.0-3.0/24)
interface GigabitEthernet0/0/0
 ip summary-address eigrp 100 10.1.0.0 255.255.252.0

Manual Summary — Named Mode

router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  af-interface GigabitEthernet0/0/0
   summary-address 10.1.0.0/22
  exit-af-interface

When a summary is configured, the router installs a Null0 route for the summary prefix and advertises only the aggregate to that interface. The summary is automatically withdrawn when no component routes remain in the topology table. Route summarisation also contains DUAL query scope, significantly improving convergence in large networks.

EIGRP Stub Routing

Stub routing is configured on spoke/leaf routers that carry no transit traffic. A stub router advertises a flag in its hello packets telling hub routers not to send DUAL queries its way — dramatically reducing query propagation and convergence time.

! Classic mode — spoke router
router eigrp 100
 eigrp stub connected summary

! Named mode — spoke router
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  eigrp stub connected summary

Stub advertisement options:

• connected — Advertise directly connected networks only
• summary — Advertise configured summary routes
• static — Advertise redistributed static routes
• redistributed — Advertise all redistributed routes
• receive-only — Accept routes; advertise nothing
• leak-map — Selectively advertise specific routes beyond stub restrictions

! Verify stub status from hub
show ip eigrp neighbors detail | include Stub

Unequal-Cost Load Balancing

EIGRP uniquely supports load balancing across unequal-cost paths using the

variance

multiplier. A Feasible Successor route is added to the routing table if its FD is within the variance multiple of the Successor FD. Traffic is distributed inversely proportional to metric.

! Classic mode
router eigrp 100
 variance 2       ! include FS paths up to 2x the best metric
 maximum-paths 4  ! up to 4 load-balanced paths in routing table

! Named mode
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  topology base
   variance 2
   maximum-paths 4
  exit-af-topology

Only Feasible Successors qualify for unequal-cost load balancing — the Feasibility Condition must still be met. A route with an RD greater than or equal to the local FD cannot be used, regardless of the variance setting.

Redistribution

Redistribute OSPF into EIGRP

router eigrp 100
 redistribute ospf 1 metric 10000 100 255 1 1500
 !                   ^BW   ^dly ^rel ^ld ^MTU

Redistribute EIGRP into OSPF

router ospf 1
 redistribute eigrp 100 subnets metric 20 metric-type E2

Redistribute Static Routes into EIGRP

router eigrp 100
 redistribute static metric 10000 100 255 1 1500

Mutual Redistribution with Route Tagging

! Tag OSPF routes redistributed into EIGRP
route-map OSPF-TO-EIGRP permit 10
 set tag 200

! Block those tags when redistributing back into OSPF (loop prevention)
route-map EIGRP-TO-OSPF deny 10
 match tag 200
route-map EIGRP-TO-OSPF permit 20

router eigrp 100
 redistribute ospf 1 metric 10000 100 255 1 1500 route-map OSPF-TO-EIGRP

router ospf 1
 redistribute eigrp 100 subnets route-map EIGRP-TO-OSPF

Always use route-maps with tag matching in mutual redistribution. Without loop prevention, a route redistributed from OSPF into EIGRP can be redistributed back into OSPF as an external route, causing routing loops and metric inflation.

Tuning Hello and Hold Timers

! Classic mode — interface level
interface GigabitEthernet0/0/0
 ip hello-interval eigrp 100 5
 ip hold-time      eigrp 100 15

! Named mode — af-interface level
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  af-interface GigabitEthernet0/0/0
   hello-interval 5
   hold-time 15
  exit-af-interface

Default values: Hello = 5 seconds (LAN/point-to-point), 60 seconds (NBMA/WAN). Hold = 3x Hello interval. Lowering timers speeds failure detection but increases bandwidth and CPU overhead on slow WAN links.

Bandwidth and Delay Tuning

! Adjust interface bandwidth — affects EIGRP metric (kbps)
interface GigabitEthernet0/0/0
 bandwidth 100000   ! 100 Mbps

! Adjust interface delay — in tens of microseconds
interface GigabitEthernet0/0/0
 delay 100          ! 1000 microseconds (100 x 10us)

Use

bandwidth

and

delay

to influence EIGRP path selection without changing K-values. Never modify K-values on a live network — all EIGRP neighbours in the AS must use identical K-values or adjacencies will not form.

Named EIGRP with IPv6

! Prerequisite
ipv6 unicast-routing

! Named EIGRP IPv6 address-family
router eigrp INFRARUNBOOK
 address-family ipv6 unicast autonomous-system 100
  af-interface GigabitEthernet0/0/0
   hello-interval 5
   hold-time 15
  exit-af-interface
  !
  topology base
  exit-af-topology
  !
  eigrp router-id 1.1.1.1
 exit-address-family

Named Mode automatically enables EIGRP on all non-passive interfaces within the address-family. A router-id must be explicitly set if the router has no IPv4 addresses — unlike Classic EIGRP for IPv6 which requires

ipv6 eigrp <AS>

on every interface individually.

EIGRP over Hub-and-Spoke

! Disable split-horizon on hub WAN interface to allow route reflection to spokes
interface GigabitEthernet0/0/0
 no ip split-horizon eigrp 100

! Named mode equivalent
router eigrp INFRARUNBOOK
 address-family ipv4 unicast autonomous-system 100
  af-interface GigabitEthernet0/0/0
   no split-horizon
  exit-af-interface

Split horizon prevents the hub from advertising routes learned from one spoke back out the same interface to other spokes. In pure hub-and-spoke topologies, disable split horizon on the hub WAN interface. If spokes have direct adjacencies between each other, leave split horizon enabled and use route summarisation at the hub instead.

Distribute Lists — Filtering EIGRP Routes

! Block 192.168.200.0/24 from being advertised out Gi0/0/1
ip prefix-list PL-BLOCK-200 seq 5  deny   192.168.200.0/24
ip prefix-list PL-BLOCK-200 seq 10 permit 0.0.0.0/0 le 32

router eigrp 100
 distribute-list prefix PL-BLOCK-200 out GigabitEthernet0/0/1

! Filter inbound routes from a specific neighbour interface
router eigrp 100
 distribute-list prefix PL-BLOCK-200 in GigabitEthernet0/0/1

Troubleshooting EIGRP

! Neighbour state
show ip eigrp neighbors
show ip eigrp neighbors detail

! Topology and path selection
show ip eigrp topology
show ip eigrp topology all-links
show ip eigrp topology summary

! Routes installed in routing table
show ip route eigrp

! Interface participation
show ip eigrp interfaces detail

! Protocol traffic counters
show ip eigrp traffic

! Event log (last 500 events)
show ip eigrp events

! Active routes (DUAL queries in progress)
show ip eigrp topology active

! Debug (use during maintenance windows only)
debug eigrp packets
debug eigrp fsm
debug eigrp neighbors

Stuck-In-Active (SIA)

SIA occurs when a DUAL query is sent and no reply arrives within the active timer (default 3 minutes per phase). The neighbour relationship is reset. Common causes:

• Congested or lossy WAN links dropping EIGRP reply packets
• High CPU on a remote router preventing DUAL reply generation
• No summarisation — queries propagate through the entire network
• No stub routing on spoke routers — hub sends queries to all neighbours

! Increase the SIA timer (default 90 seconds per active-timer phase)
router eigrp 100
 timers active-time 120

! Identify SIA routes
show ip eigrp topology active

! Identify which neighbour is not replying
show ip eigrp neighbors detail | include SIA

The correct long-term fix for SIA is route summarisation and stub routing — not simply increasing the active timer. Summarisation contains the query domain; stub routing prevents queries from reaching endpoint routers entirely.

EIGRP Verification Checklist

• show ip eigrp neighbors

  — Adjacencies up, Q Cnt = 0 on all neighbours

• show ip eigrp topology

  — Expected routes present in Passive state with Successors

• show ip eigrp topology all-links

  — Feasible Successors exist for critical routes

• show ip route eigrp

  — EIGRP routes installed correctly in RIB

• show ip eigrp interfaces

  — Correct interfaces participating, passive where expected

• show ip eigrp traffic

  — No excessive query/reply traffic indicating instability

• show ip eigrp events

  — No repeated neighbour state changes

---

Frequently Asked Questions

What is the difference between Classic EIGRP and Named EIGRP?

Classic EIGRP uses

router eigrp <AS-number>

with separate IPv4 and IPv6 processes; interface-specific settings are configured under the interface itself. Named EIGRP uses

router eigrp <name>

with address-family stanzas, supporting IPv4, IPv6, and VRF within a single process with centralised per-interface configuration under

af-interface

blocks. Named mode also supports SHA-256 authentication and is the current standard for IOS-XE deployments.

Why will my EIGRP neighbours not form?

Common causes: mismatched AS numbers between neighbours, mismatched K-values, authentication mismatch (wrong key or algorithm),

passive-interface

configured on the peering interface, no

network

statement covering the interface IP address, or an ACL blocking multicast 224.0.0.10. Run

debug eigrp packets

and check

show ip eigrp neighbors detail

to isolate the cause. Confirm the interface is participating with

show ip eigrp interfaces

.

What is the Feasibility Condition?

A backup route qualifies as a Feasible Successor if its Reported Distance (the metric the neighbour advertises for the route) is strictly less than the local router current Feasible Distance. This mathematical guarantee ensures the backup path cannot create a routing loop — the neighbour must be closer to the destination than the local router currently believes it is. Routes that do not meet the Feasibility Condition still appear in the topology table via

show ip eigrp topology all-links

but are not usable as immediate backups.

Does EIGRP support unequal-cost load balancing?

Yes — EIGRP is the only standard routing protocol that natively supports unequal-cost load balancing. Use the

variance

command with a multiplier value. Any Feasible Successor route with an FD within (multiplier x Successor FD) is added to the routing table. Traffic is distributed proportionally to the inverse of each path metric — paths with lower metrics carry more traffic. Only Feasible Successors qualify; routes that fail the Feasibility Condition cannot participate regardless of variance.

What metric values should I use when redistributing into EIGRP?

You must provide a seed metric when redistributing from non-EIGRP sources. The format is

metric bandwidth delay reliability load mtu

where bandwidth is in kbps, delay is in tens of microseconds, reliability and load are 0-255, and MTU is in bytes. A commonly used safe default is

metric 10000 100 255 1 1500

— representing a 10 Mbps link with 1 ms delay, full reliability, minimal load, and standard MTU. Without a seed metric, redistributed routes have an infinite metric and will not be installed.

When should I use EIGRP Stub routing?

Configure stub routing on any router that is purely an endpoint — not a transit path for other networks. Spoke routers in hub-and-spoke topologies, branch office routers with a single uplink, and distribution-layer switches with no alternate paths are ideal candidates. A stub router advertises a flag in its hello packets that tells hub routers not to send DUAL queries to it. This contains the query domain, reduces convergence time, and lowers CPU load on hub routers during topology changes.

How does EIGRP summarisation differ from OSPF summarisation?

EIGRP summarisation is performed per-interface on any router in the topology — there is no area boundary requirement as in OSPF. A summary can be configured outbound on any EIGRP-facing interface. The summarising router installs a Null0 discard route for the summary prefix and advertises only the aggregate to that interface. The summary is withdrawn automatically when no component routes remain in the topology table. Summarisation also contains the DUAL query domain, which is a key convergence benefit in large networks.

What causes Stuck-In-Active (SIA) and how should I resolve it?

SIA occurs when DUAL sends a query to a neighbour and no reply is received within the active timer (approximately 3 minutes). The neighbour adjacency is reset. Root causes include congested links dropping EIGRP packets, high CPU preventing reply generation, and overly broad query scope due to missing summarisation or stub routing. The correct resolution is route summarisation to limit query propagation and stub routing on spoke routers. Increasing the active timer with

timers active-time

is a temporary measure — it masks the underlying topology design issue.

Can I run multiple EIGRP AS numbers on one router?

Yes. Multiple

router eigrp <AS>

instances can coexist on a single router, each with its own independent topology table, neighbour table, and routing process. Routes do not leak automatically between EIGRP AS instances — redistribution is required at boundary routers. This is used in network designs that connect separate EIGRP administrative domains. Named Mode handles multiple address families and VRFs within a single named process, eliminating the need for separate AS instances in most modern designs.

How does EIGRP convergence compare to OSPF?

EIGRP converges faster than OSPF when a Feasible Successor exists for the failed route — the router simply promotes the FS to Successor with no recalculation or neighbour queries. If no FS exists, EIGRP enters Active state and convergence depends on query propagation depth and reply latency. OSPF always runs an SPF calculation on topology change but uses incremental SPF to limit scope. In well-designed EIGRP networks with summarisation, stub routing, and adequate Feasible Successors, convergence is typically sub-second. In flat EIGRP networks without summarisation, SIA events can cause convergence times exceeding several minutes.

What is the EIGRP router-id used for?

The EIGRP router-id is a 32-bit dotted-decimal value that identifies the originator of external (redistributed) routes in the EIGRP topology table. It is used to prevent a router from accepting its own redistributed routes back as external entries — a loop-prevention mechanism. Unlike OSPF, the EIGRP router-id has no effect on neighbour selection or path calculation. Best practice is to set it explicitly using

eigrp router-id <loopback-ip>

to ensure it remains stable across interface state changes.

How do I influence EIGRP path selection without changing K-values?

Adjust the

bandwidth

and

delay

interface parameters. Bandwidth is in kbps and delay is in tens of microseconds — both directly affect the EIGRP composite metric. Raise the delay or lower the bandwidth on a less-preferred path to increase its metric. Alternatively, use offset-lists to add a fixed metric increment to specific routes on specific interfaces:

offset-list <acl> in|out <offset> <interface>

under the router eigrp process. This approach is more surgical than bandwidth or delay changes and does not affect other protocols that reference those interface parameters.

How do I verify that route summarisation is working correctly?

Run

show ip route eigrp

and confirm the summary route appears with a Null0 egress interface on the summarising router. On downstream neighbours, run

show ip eigrp topology

and verify only the summary prefix is received — not the individual component routes. If component routes appear alongside the summary, verify the

ip summary-address eigrp

command is applied on the correct outbound interface. Confirm with

show ip eigrp interfaces detail

that the summary is listed under the correct interface output.