EIGRP (Enhanced Interior Gateway Routing Protocol) is a Cisco-proprietary distance vector routing protocol that combines the best features of distance vector and link-state protocols. For the CCNA 200-301 exam (objective 3.5), you must understand EIGRP's unique operation, including its Diffusing Update Algorithm (DUAL), feasible successors, and metric calculation. Mastering EIGRP is essential because it is widely deployed in enterprise networks for its fast convergence and low overhead.
Jump to a section
Imagine you are driving to a destination using a GPS navigation system. Your GPS knows the primary route (the successor) and also keeps a list of alternate routes (feasible successors) that are guaranteed not to loop back to your starting point. The GPS continuously checks road conditions; if the primary route becomes blocked (link failure), it instantly switches to the best alternate route without recalculating from scratch. This is exactly how EIGRP works: each router maintains a topology table with all known routes to a destination, and the best route is called the successor. For each destination, there may be one or more feasible successors—neighbors that have a reported distance less than the feasible distance. If the successor fails, EIGRP uses DUAL to immediately promote the best feasible successor to successor, providing sub-second convergence. The GPS analogy breaks down slightly because EIGRP uses a metric (composite of bandwidth, delay, load, reliability) to determine the best path, whereas GPS uses distance and traffic. However, the core idea of having pre-computed backup routes is identical. In EIGRP, the feasible successor condition ensures that the alternate path does not create a routing loop, much like how a GPS only suggests an alternate route if it knows it leads away from the original blocked segment.
What is EIGRP and Why It Exists
EIGRP is a Cisco-proprietary routing protocol introduced in 1992 as an enhancement to IGRP (Interior Gateway Routing Protocol). It is a hybrid protocol: it behaves like a distance vector protocol in that it advertises routes with metrics and uses the Bellman-Ford algorithm (though DUAL is a loop-free distance vector algorithm), but it also maintains neighbor relationships and uses hello packets like a link-state protocol. EIGRP was designed to overcome the slow convergence and potential loops of traditional distance vector protocols like RIP. It is classless (supports VLSM and CIDR), supports authentication, and can load-balance across unequal-cost paths (via variance). EIGRP is widely used in Cisco-only environments because of its fast convergence, low overhead, and ease of configuration.
How EIGRP Works Step by Step
EIGRP operation can be broken down into three phases: neighbor discovery, topology exchange, and route selection.
Neighbor Discovery: - Routers send Hello packets every 5 seconds (default) on high-bandwidth links (>= 1.544 Mbps) or every 60 seconds on low-speed links (<= 1.544 Mbps). - Hello packets are multicast to 224.0.0.10. - A neighbor is considered established when a Hello is received with matching AS number, K-values, and authentication parameters. - Hold time is 3 times the Hello interval (15 seconds default on high-speed, 180 on low-speed). If no Hello is received within the hold time, the neighbor is declared dead.
Topology Exchange: - Once a neighbor is established, routers exchange full routing tables via Update packets (reliable multicast). - Updates are sent only when a route changes, not periodically (unlike RIP). - EIGRP uses Reliable Transport Protocol (RTP) to ensure reliable delivery of Update, Query, and Reply packets. Hello and ACK packets are unreliable. - The topology table stores all learned routes, including the feasible distance (FD) – the metric of the best path to a destination – and the reported distance (RD) from each neighbor.
Route Selection with DUAL: - DUAL (Diffusing Update Algorithm) is used to select loop-free paths. - For each destination, the router selects the neighbor with the lowest metric as the successor (primary route). - A feasible successor is a neighbor whose reported distance (RD) is less than the current feasible distance (FD). This ensures that the alternate path does not create a routing loop (the feasibility condition). - If a successor fails and a feasible successor exists, the router immediately promotes the best feasible successor to successor. No queries are sent. - If no feasible successor exists, the router enters Active state for that route and sends Query packets to all neighbors. Neighbors must reply with their own distances. This process can cause a delay in convergence.
Key States, Timers, and Defaults
Neighbor States: Init, Pending, Up, Down (simplified; EIGRP uses a finite state machine with states like 1-Way, 2-Way, etc.)
Hello Interval: 5 seconds (default for most interfaces); 60 seconds for NBMA interfaces (e.g., Frame Relay multipoint).
Hold Time: 15 seconds (default for most interfaces); 180 seconds for NBMA.
Active Timer: 3 minutes (default). If a query is not answered within this time, the route is stuck-in-active (SIA) and the neighbor is reset.
Metric Weights (K-values): K1=1 (bandwidth), K2=0 (load), K3=1 (delay), K4=0 (reliability), K5=0. Only bandwidth and delay are used by default.
Metric Calculation: Metric = (K1 * bandwidth + (K2 * bandwidth)/(256 - load) + K3 * delay) * 256. With defaults: Metric = (bandwidth + delay) * 256. Bandwidth is 10^7 / minimum bandwidth (in kbps) along the path. Delay is sum of delays (in tens of microseconds) divided by 10.
Variance: Default 1 (equal-cost load balancing only). Can be set to allow unequal-cost load balancing.
IOS CLI Verification Commands
Show EIGRP Neighbors:
R1# show ip eigrp neighbors
EIGRP-IPv4 Neighbors for AS(100)
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
0 10.1.1.2 GigabitEthernet0/0 13 00:12:34 1 100 0 12H: Handle number (locally assigned)
Hold: Current hold time remaining (seconds)
Uptime: Time since neighbor was first discovered
SRTT: Smooth Round-Trip Time (milliseconds)
RTO: Retransmission Timeout (milliseconds)
Q Cnt: Queue count (should be 0)
Seq Num: Sequence number of last packet received
Show IP EIGRP Topology:
R1# show ip eigrp topology
EIGRP-IPv4 Topology Table for AS(100)
Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,
r - reply Status, s - sia Status
P 192.168.1.0/24, 1 successors, FD is 28160
via 10.1.1.2 (28160/28160), GigabitEthernet0/0
via 10.1.2.2 (30720/28160), GigabitEthernet0/1FD: Feasible Distance (metric of best path)
The numbers in parentheses are (computed metric / reported distance)
'P' means Passive (stable). 'A' means Active (query in progress).
Show IP Protocols:
R1# show ip protocols
*** IP Routing is NSF aware ***
Routing Protocol is "eigrp 100"
Outgoing update filter list for all interfaces is not set
Incoming update filter list for all interfaces is not set
Default networks flagged in outgoing updates
Default networks accepted from incoming updates
EIGRP-IPv4 Protocol for AS(100)
Metric weight K1=1, K2=0, K3=1, K4=0, K5=0
NSF-aware route hold timer is 240
Router-ID: 1.1.1.1
Topology : 0 (base)
Active Timer: 3 min
Distance: internal 90 external 170
Maximum path: 4
Maximum hopcount 100
Maximum metric variance 1How EIGRP Interacts with Related Protocols
EIGRP uses protocol number 88 (IP). It does not rely on TCP or UDP; instead, it uses its own Reliable Transport Protocol (RTP) for reliable delivery of updates, queries, and replies. EIGRP can redistribute routes from other routing protocols (RIP, OSPF, static) and can be configured with route maps to filter routes. EIGRP also supports route summarization and can be used with MPLS VPNs. On the CCNA exam, you must know that EIGRP has an administrative distance of 90 for internal routes and 170 for external routes.
Configure EIGRP on Routers
First, enable EIGRP globally with the router eigrp command followed by an autonomous system (AS) number. All routers in the same EIGRP domain must use the same AS number. Then, use the network command to advertise directly connected networks. The network command uses a wildcard mask (inverse of subnet mask) to specify which interfaces participate. Example: ``` R1(config)# router eigrp 100 R1(config-router)# network 10.0.0.0 0.255.255.255 R1(config-router)# network 192.168.1.0 0.0.0.255 ``` This enables EIGRP on interfaces with IP addresses in those ranges. Alternatively, you can use the network command without a wildcard mask to advertise major classful networks (not recommended).
Verify Neighbor Adjacency
After configuration, verify that routers have formed neighbor adjacencies. Use show ip eigrp neighbors to see the neighbor table. Ensure that the Hold column shows a value less than the hold time (default 15 seconds) and that the Q Cnt is 0. If no neighbors appear, check that interfaces are up/up, IP addresses are in the same subnet, the AS number matches, and there are no ACLs blocking multicast 224.0.0.10. Example: ``` R1# show ip eigrp neighbors EIGRP-IPv4 Neighbors for AS(100) H Address Interface Hold Uptime SRTT RTO Q Seq (sec) (ms) Cnt Num 0 10.1.1.2 GigabitEthernet0/0 13 00:12:34 1 100 0 12 ```
Examine the Topology Table
Use show ip eigrp topology to view all routes learned by EIGRP. Look for routes marked 'P' (Passive) – these are stable. If a route is 'A' (Active), there is a query in progress. The topology table shows the feasible distance (FD) and the reported distance (RD) from each neighbor. Identify the successor (the route with the lowest FD) and any feasible successors (routes where RD < FD). Example: ``` R1# show ip eigrp topology P 192.168.1.0/24, 1 successors, FD is 28160 via 10.1.1.2 (28160/28160), GigabitEthernet0/0 via 10.1.2.2 (30720/28160), GigabitEthernet0/1 ```
Check the Routing Table
EIGRP routes are installed in the IP routing table with an administrative distance of 90 (internal). Use show ip route eigrp to display only EIGRP-learned routes. Verify that the successor route appears with the correct next-hop and metric. Example: ``` R1# show ip route eigrp 192.168.1.0/24 is variably subnetted, 2 subnets, 2 masks D 192.168.1.0/24 [90/28160] via 10.1.1.2, 00:12:34, GigabitEthernet0/0 ``` The 'D' indicates EIGRP (Distance Vector). The bracket shows [administrative distance/metric].
Tune Metric or Load Balancing
By default, EIGRP performs equal-cost load balancing (up to 4 paths). To enable unequal-cost load balancing, use the variance command under router eigrp. Variance multiplies the FD; any feasible successor with a metric less than variance * FD will be used. Example: ``` R1(config-router)# variance 2 ``` This allows routes with metric up to 2 times the FD to be installed. You can also adjust the metric by changing delay or bandwidth on an interface (though changing bandwidth can affect QoS). Use the interface command delay and bandwidth to influence EIGRP metric. However, be cautious: changing bandwidth affects other protocols like OSPF.
Troubleshoot EIGRP Issues
Common issues include neighbor flapping (due to mismatched K-values or AS numbers) and stuck-in-active (SIA) routes. Use debug eigrp packets (with caution) to see Hello and Update packets. For SIA, check show ip eigrp topology active to see which routes are Active and which neighbors have not replied. Increase the active timer if necessary, but the root cause is usually a slow or lossy link. Example: ``` R1# show ip eigrp topology active A 192.168.2.0/24, 1 successors, FD is 28160 via 10.1.1.2 (28160/28160), GigabitEthernet0/0, serno 12 via 10.1.2.2 (30720/28160), GigabitEthernet0/1, serno 14 ``` This shows the route is Active (query in progress).
In enterprise networks, EIGRP is often used as the internal routing protocol within a data center or campus network. For example, consider a three-tier architecture with core, distribution, and access layers. EIGRP can be configured on all routers to provide fast convergence in case of link failures. A typical deployment might use EIGRP with a single AS number (e.g., 100) and use passive interfaces on access-layer routers to prevent unnecessary neighbor relationships with end devices. The network engineer would configure summarization at the distribution layer to reduce the size of routing tables and minimize query propagation. For instance, a distribution router might summarize a range of /24 subnets into a single /16 route.
Another common scenario is using EIGRP for load balancing across multiple WAN links. With the variance command, engineers can take advantage of unequal bandwidth links (e.g., a 100 Mbps fiber and a 50 Mbps DSL) by adjusting the metric so that traffic is distributed proportionally. The engineer would first ensure that the slower link is a feasible successor, then set variance to allow its metric. This is more efficient than equal-cost load balancing, which would send 50% of traffic over each link regardless of capacity.
When misconfigured, EIGRP can cause routing loops or black holes. For example, if two routers are configured with different K-values, they will not become neighbors. Or, if an access list blocks multicast 224.0.0.10, neighbor adjacency will fail. In production, a common mistake is to forget to include a network statement for a new subnet, causing that subnet to be unreachable. Troubleshooting involves checking the neighbor table, topology table, and routing table step by step. EIGRP's scalability is excellent; it can handle hundreds of routers in a single AS, but careful summarization is needed to avoid excessive query traffic during convergence.
The CCNA 200-301 exam objective 3.5 covers EIGRP configuration, verification, and troubleshooting. Specifically, you need to know how to configure EIGRP (router eigrp AS, network statements), verify neighbor relationships (show ip eigrp neighbors), and interpret the topology table (show ip eigrp topology). You should also understand DUAL and the concept of feasible successors.
Common wrong answers on the exam: 1. Believing EIGRP is a link-state protocol. Many candidates confuse EIGRP's use of Hello packets and neighbor tables with OSPF. Remember: EIGRP is a distance vector protocol (advanced distance vector). 2. Thinking that EIGRP uses TCP or UDP. EIGRP uses its own Reliable Transport Protocol (RTP) on top of IP (protocol 88). 3. Mixing up the metric formula. Candidates often forget that by default only bandwidth and delay are used (K1=1, K3=1). Load and reliability are not used unless K2 or K4/K5 are non-zero. 4. Assuming that the successor is the route with the lowest reported distance. Actually, the successor is the neighbor with the lowest computed metric (FD). The reported distance (RD) is used for feasibility condition.
Specific values to memorize: AD 90 (internal), 170 (external); default Hello interval 5 seconds (high-speed), 60 seconds (low-speed); hold time 15 seconds (high-speed), 180 seconds (low-speed); multicast address 224.0.0.10; maximum paths default 4.
Calculation trap: When computing the EIGRP metric, remember bandwidth is 10^7 / min bandwidth (in kbps) along the path. Delay is sum of interface delays (in tens of microseconds) divided by 10. For example, if a path has min bandwidth 100 Mbps (100,000 kbps) and total delay 1000 microseconds (100 tens of microseconds), metric = (10^7/100000 + 100/10) * 256 = (100 + 10) * 256 = 28160.
Decision rule: If a question asks which path EIGRP will choose, compute the metric for each path. The path with the lowest metric wins. If two paths have the same metric, they are installed (equal-cost load balancing). If a feasible successor exists and the successor fails, convergence is immediate.
EIGRP is an advanced distance vector protocol using DUAL for loop-free fast convergence.
Neighbor adjacency uses multicast 224.0.0.10 with Hello interval 5 sec (default high-speed) and hold time 15 sec.
EIGRP metric by default uses bandwidth (K1=1) and delay (K3=1); metric = (bandwidth + delay) * 256.
Feasible successor condition: RD < FD. If successor fails, feasible successor is immediately promoted.
Administrative distance: internal 90, external 170.
Use show ip eigrp neighbors, show ip eigrp topology, show ip route eigrp for verification.
EIGRP uses its own Reliable Transport Protocol (RTP) over IP protocol 88.
These come up on the exam all the time. Here's how to tell them apart.
EIGRP
Cisco proprietary (though partially open via RFC 7868)
Uses DUAL algorithm for loop-free convergence
Metric based on bandwidth and delay (default)
Supports unequal-cost load balancing (variance)
Administrative distance: 90 internal, 170 external
OSPF
Open standard (RFC 2328)
Uses SPF algorithm (Dijkstra) for loop-free paths
Metric based on cost (reference bandwidth / interface bandwidth)
Only equal-cost load balancing by default
Administrative distance: 110
Mistake
EIGRP is a link-state routing protocol.
Correct
EIGRP is an advanced distance vector (hybrid) protocol. It uses distance vector logic (Bellman-Ford/DUAL) but has link-state features like neighbor discovery and hello packets.
Because EIGRP forms neighbor relationships and uses hello packets, it appears similar to OSPF (a link-state protocol).
Mistake
EIGRP uses TCP or UDP for reliable delivery.
Correct
EIGRP uses its own Reliable Transport Protocol (RTP) directly over IP (protocol 88). It does not use TCP or UDP.
Many candidates assume all reliable transport uses TCP or UDP, but EIGRP has its own reliability mechanism.
Mistake
The successor is the route with the lowest reported distance from any neighbor.
Correct
The successor is the neighbor with the lowest computed metric (feasible distance). The reported distance is used only to determine feasible successors (if RD < FD).
Candidates often confuse reported distance with feasible distance.
Mistake
EIGRP uses all five K-values by default.
Correct
By default, only K1 (bandwidth) and K3 (delay) are set to 1. K2, K4, and K5 are 0, meaning load and reliability are not used in metric calculation.
Because the K-values are configurable, some assume they are all active by default.
Reveal each answer, then mark whether you got it right. Score 60%+ to unlock the next chapter.
EIGRP is a Cisco-proprietary advanced distance vector protocol using DUAL, while OSPF is an open-standard link-state protocol using SPF. EIGRP uses a composite metric (bandwidth and delay by default) and supports unequal-cost load balancing. OSPF uses cost (based on bandwidth) and only equal-cost load balancing. EIGRP has administrative distance 90 (internal), OSPF has 110. On the exam, know that EIGRP converges faster because of feasible successors, but OSPF is more scalable in large networks.
Enter global configuration mode and use 'router eigrp <AS-number>'. Then use 'network <network> [wildcard-mask]' to advertise interfaces. For example: 'router eigrp 100' then 'network 10.0.0.0 0.255.255.255'. All routers must use the same AS number. You can also use passive-interface to suppress hellos on interfaces that don't need neighbors.
Stuck-in-active (SIA) occurs when a router sends a Query for a route and does not receive a Reply from a neighbor within the active timer (default 3 minutes). This usually indicates a network problem (e.g., packet loss, high delay). The router then resets the neighbor relationship, causing route flapping. To avoid SIA, ensure good network connectivity and use summarization to limit query scope.
Yes, using the 'variance' command. By default, variance is 1, meaning only equal-cost paths are used. To allow unequal-cost load balancing, set variance to a multiplier (e.g., variance 2 allows paths with metric up to 2 times the best metric). The alternate path must be a feasible successor. Example: 'router eigrp 100' then 'variance 2'.
The default metric uses bandwidth and delay: Metric = (bandwidth + delay) * 256, where bandwidth = 10^7 / minimum bandwidth (in kbps) along the path, and delay = sum of interface delays (in tens of microseconds) / 10. For example, if min bandwidth = 100000 kbps and total delay = 100 tens of microseconds, metric = (100 + 10) * 256 = 28160.
Key commands: 'show ip eigrp neighbors' (neighbor table), 'show ip eigrp topology' (topology table), 'show ip route eigrp' (routing table), 'show ip protocols' (EIGRP configuration). Use 'debug eigrp packets' for troubleshooting (with caution).
Hello packets are used to discover and maintain neighbor relationships. They are sent every 5 seconds (default) on most interfaces to multicast address 224.0.0.10. If a router does not receive a hello from a neighbor within the hold time (default 15 seconds), it declares the neighbor down. Hello packets are unreliable (no ACK required).
You've just covered EIGRP Overview — now see how well it sticks with free CCNA 200-301 practice questions. Full explanations included, no account needed.
Done with this chapter?