10/23/2012

EIGRP

• “Enhanced” Interior Gateway Routing Protocol

Chapter 4/Module 3 • Based on IGRP and developed to allow easy transition from IGRP to
EIGRP. (“Like IGRP+”)
EIGRP
• Cisco proprietary, released in 1994
• EIGRP is an advanced distance-vector routing protocol that relies on
features commonly associated with link-state protocols. (sometimes
CCNA 3 version 3.0 called a hybrid routing protocol).
• Note: The Hybrid term sometimes misleads people into thinking
EIGRP has the topology benefits of a link state routing protocol. It
does not. EIGRP is a distance vector routing protocol and suffers from
all of the same disadvantages of any other distance vector routing
protocol, i.e. routing loops.
• Note: “Often described as a hybrid routing protocol offering the best of
distance-vector and link-state algorithms.” - I would say “features of
distance-vector and link-state” not necessarily “the best.”

IGRP and EIGRP: A migration path Metric Calculation (Review)

IGRP EIGRP

Classful Routing Protocol Classless Routing Protocol

• VLSM, CIDR

bandwidth = (10,000,000/bandwidth kbps) bandwidth = (10,000,000/bandwidth kbps) * 256

delay = delay/10 delay = (delay/10) * 256
24 bit metric for bandwidth and delay 32 bit metric for bandwidth and delay
Maximum Hop Count = 255 Maximum Hop Count = 224

No differentiation between internal and Outside routes (redistributed) are tagged as

external routes. external routes(AD=170). – k1 for bandwidth
AD for EIGRP Internal Routes=90 – k2 for load
Automatic redistribution between IGRP and EIGRP as long as “AS” numbers are the same. – k3 for delay
– k4 and k5 for Reliability
EIGRP
Router(config-router)# metric
bandwidth is in kbps weights tos k1 k2 k3 k4 k5

1
 10/23/2012

Displaying Interface Values Displaying Interface Values

Router> show interface s0/0 Router> show interface s0/0

Serial0/0 is up, line protocol is up Serial0/0 is up, line protocol is up
Hardware is QUICC Serial Bandwidth Delay Hardware is QUICC Serial Bandwidth Delay
Description: Out to VERIO Description: Out to VERIO
Internet address is 207.21.113.186/30 Internet address is 207.21.113.186/30
MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec, MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
rely 255/255, load 246/255 rely 255/255, load 246/255
Encapsulation PPP, loopback not set Encapsulation PPP, loopback not set
Keepalive set (10 sec) Keepalive set (10 sec)
<output omitted> Reliability Load <output omitted> Reliability Load

shows reliability as a fraction of 255, for shows load as a fraction of 255, for
example (higher is better): example (lower is better):
Routing Table Metric
rely 190/255 (or 74% reliability) load 10/255 (or 3% loaded link) • Default: Slowest of bandwidth plus the sum of the delays of
rely 234/255 (or 92% reliability) load 40/255 (or 16% loaded link) all outgoing interfaces from “this router” to the destination
rely 255/255 (or 100% reliability) load 255/255 (or 100% loaded link) network.

EIGRP Metrics
Values displayed in show interface SanJose2#show ip route
commands and sent in routing updates.
D 192.168.72.0/24 [90/2172416]
Bandwidth BWEIGRP DLYEIGRP
Delay/10
via 192.168.64.6, 00:28:26, Serial0
Media K= kilobits 10,000,000/Bandwidth Delay
*256
*256
100M ATM 100,000K 25,600 100 S 2,560
Fast Ethernet 100,000K 25,600 100 S 2,560
FDDI 100,000K 25,600 100 S 2,560
HSSI 45,045K 56,832 20,000 S 512,000
16M Token Ring 16,000K 160,000 630 S 16,128
A Closer Look at the Routing Table
Ethernet 10,000K 256,000 1,000 S 25,600
T1 (Serial 1,544K 1,657,856 20,000 S 512,000 Metrics
Default)
512K 512K 4,999,936 20,000 S 512,000
DS0 64K 40,000,000 20,000 S 512,000
56K 56K 45,714,176 20,000 S 512,000

BWEIGRP and DLYEIGRP display values as sent in EIGRP updates and used in
calculating the EIGRP metric. Calculated values (cumulative) displayed in
routing table (show ip route).

2
 10/23/2012

The Routing Table Displaying Interface Values

How does SanJose2 calculate the cost for this route?
Westasman> show interface fa0/0
Ethernet0 is up, line protocol is up
Hardware is Lance, address is 0010.7b3a.cf84 (bia 0010.7b3a.cf84)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
rely 255/255, load 1/255
<output omitted>

SanJose2> show interface s0/0

Serial0/0 is up, line protocol is up
Hardware is QUICC Serial
Description: Out to Westasman
Internet address is 192.168.64.5/30
MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
rely 255/255, load 246/255
SanJose2#show ip route Administrative Distance / Metric <output omitted>

D 192.168.72.0/24 [90/2172416]
via 192.168.64.6, 00:28:26, Serial0

Determining the costs Determining the costs

Bandwidth = (10,000,000/bandwidth kbps) * 256 Delay = (delay/10) * 256
Delay = Delay =
FastEthernet 2,560 FastEthernet 2,560
= (10,000,000/100,000) * 256 Fa0/0 192.168.72.1/24 = (100/10) * 256 Fa0/0 192.168.72.1/24
= 25,600 Bandwidth = = 2,560 Bandwidth =
Westasman
25,600 Westasman
25,600
S0/0 192.168.64.2/30 S0/1 192.168.64.6/30 S0/0 192.168.64.2/30 S0/1 192.168.64.6/30
T1 T1
= (10,000,000/1544) * 256 Delay = = (20,000/10) * 256 Delay =
= 1,657,856 512,000 = 512,000 512,000

S0/0 192.168.64.1/30 S0/0 192.168.64.5/30 S0/0 192.168.64.1/30 S0/0 192.168.64.5/30

Fa0/0 192.168.1.2/24 Bandwidth = Fa0/0 192.168.1.2/24 Bandwidth =
SanJose1
1,657,856 SanJose1
1,657,856
SanJose2 SanJose2

Fa0/0 192.168.1.1/24 Fa0/0 192.168.1.1/24

EIGRP EIGRP
AS 100 AS 100

3
 10/23/2012

Determining the costs The Routing Table

What is the cost (metric) for 192.168.72.0/24 from SanJose2?
Delay =
Cost: Slowest bandwidth
2,560
+ sum of delays Fa0/0 192.168.72.1/24
1,657,856 Bandwidth =
512,000 Westasman
25,600
2,560 S0/0 192.168.64.2/30 S0/1 192.168.64.6/30
--------------
2,172,416
The cost! Delay =
512,000

S0/0 192.168.64.1/30 S0/0 192.168.64.5/30

Fa0/0 192.168.1.2/24 Bandwidth =
SanJose1
1,657,856
SanJose2

Fa0/0 192.168.1.1/24 Slowest!

SanJose2#show ip route Administrative Distance / Metric
bandwidth = (10,000,000/bandwidth kbps) * 256
EIGRP D 192.168.72.0/24 [90/2172416]
delay = (delay/10) * 256
AS 100 via 192.168.64.6, 00:28:26, Serial0

EIGRP and IGRP compatibility EIGRP and IGRP compatibility

External

External

10,476 = 6,476(BW)+2,000(DLY)+2,000(DLY)
IGRP Metrics! (Does not multiply by 256.
• Automatic redistribution occurs when the same AS number is used for
EIGRP and IGRP. • EIGRP will tag routes learned from IGRP, or any outside source, as
• EIGRP scales the IGRP metric by a factor of 256. external because they did not originate from EIGRP routers.
• IGRP reduces the metric by a factor of 256. • IGRP cannot differentiate between internal and external routes.

4
 10/23/2012

Features of EIGRP EIGRP Terminology

• Classless Routing Protocol (VLSM, CIDR) • Neighbor table – Each EIGRP router maintains a neighbor table that
lists adjacent routers. This table is comparable to the adjacency
• Faster convergence times and improved scalability database used by OSPF. There is a neighbor table for each protocol
• Multiprotocol support: TCP/IP, IPX/SPX, Appletalk that EIGRP supports.
– There is no IPX/SPX or Appletalk in CCNA or CCNP • Topology table – Every EIGRP router maintains a topology table for
• Rapid Convergence and Better handling of routing loops – (DUAL) (coming) each configured network protocol. This table includes route entries for
• Efficient Use of Bandwidth all destinations that the router has learned. All learned routes to a
destination are maintained in the topology table.
– Partial, bounded updates: Incremental updates only to the routers that need
them. • Routing table – EIGRP chooses the best routes to a destination from
the topology table and places these routes in the routing table. Each
– Minimal bandwidth consumption: Uses Hello packets and EIGRP packets
EIGRP router maintains a routing table for each network protocol.
by default use no more than 50% of link’s bandwidth EIGRP packets.
• Successor – A successor is a route selected as the primary route to
• PDM (Protocol Dependent Module) use to reach a destination. Successors are the entries kept in the
– Keeps EIGRP modular routing table. Multiple successors for a destination can be retained in
– Different PDMs can be added to EIGRP as new routed protocols are the routing table.
enhanced or developed: IPv4, IPv6, IPX, and AppleTalk • Feasible successor – A feasible successor is a backup route. These
• Unequal-cost load balancing same as IGRP (unlike OSPF) routes are selected at the same time the successors are identified, but
are kept in the topology table. Multiple feasible successors for a
destination can be retained in the topology table.

Neighbor Table Neighbor Table

• Each EIGRP router maintains a neighbor table that lists adjacent RouterC#show ip eigrp neighbors
routers. IP-EIGRP neighbors for process 44
• This table is comparable to the adjacency database used by OSPF. H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
• There is a neighbor table for each protocol that EIGRP supports
0 192.168.0.1 Se0 11 00:03:09 1138 5000 0 6
• Whenever a new neighbor is discovered, the address of that neighbor 1 192.168.1.2 Et0 12 00:34:46 4 200 0 4
and the interface used to reach it are recorded in a new neighbor table
entry.
RouterC#show ip eigrp neighbors • Neighbor address The network-layer address of the neighbor
IP-EIGRP neighbors for process 44
router(s).
H Address Interface Hold Uptime SRTT RTO Q Seq • Queue count The number of packets waiting in queue to be sent. If
(sec) (ms) Cnt Num
this value is constantly higher than zero, then there may be a
congestion problem at the router. A zero means that there are no
0 192.168.0.1 Se0 11 00:03:09 1138 5000 0 6
EIGRP packets in the queue.
1 192.168.1.2 Et0 12 00:34:46 4 200 0 4

5
 10/23/2012

Neighbor Table Neighbor Table

• Note that an EIGRP router can maintain multiple neighbor tables, one for each L3
RouterC#show ip eigrp neighbors protocol running (for example, IP, AppleTalk).
IP-EIGRP neighbors for process 44 • A router must run a unique EIGRP process for each routed protocol.
H Address Interface Hold Uptime SRTT RTO Q Seq RTX#show ip eigrp neighbors
(sec) (ms) Cnt Num IP-EIGRP neighbors for process 1
0 192.168.0.1 Se0 11 00:03:09 1138 5000 0 6 H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
1 192.168.1.2 Et0 12 00:34:46 4 200 0 4
1 10.2.0.2 Se1 12 00:27:39 333 1998 0 10
0 10.1.0.1 Se0 14 01:17:14 40 240 0 27

• Smooth Round Trip Timer (SRTT) The average time it takes to send RTX#show ipx eigrp neighbors
and receive packets from a neighbor. IPX EIGRP Neighbors for process 22
H Address Interface Hold Uptime SRTT RTO Q Seq
• This timer is used to determine the retransmit interval (RTO) (sec) (ms) Cnt Num
1 2000.0000.0c76.080c Se1 14 00:04:21 28 200 0 22
• Hold Time The interval to wait without receiving anything from a 0 1000.0000.0c38.6fa2 Se0 14 00:04:24 28 200 0 22
neighbor before considering the link unavailable. RTX#show appletalk eigrp neighbors
• Originally, the expected packet was a hello packet, but in current AT/EIGRP Neighbors for process 1, router id 2
Cisco IOS software releases, any EIGRP packets received after H Address Interface Hold Uptime SRTT RTO Q Seq
the first hello will reset the timer. (sec) (ms) Cnt Num
0 1000.123 Se0 11 00:15:01 8 200 0 7
1 2000.28 Se1 14 00:41:11 11 200 0 9

Topology Table Topology Table – Extra Information

• Topology table • Not only does the topology table track information regarding route states, but it
– Each EIGRP router maintains a topology table for each configured can also record special information for external routes, including the
network protocol. administrator tag.
– This table includes route entries for all destinations that the router has • EIGRP classifies routes as either internal or external.
learned. • EIGRP uses a process called route tagging to add special tags to each route.
– All learned routes to a destination are maintained in the topology table. • These tags identify a route as internal or external, and may include other
information as well.
• EIGRP uses its topology table to store all the information it needs to
calculate a set of distances and vectors to all reachable destinations.
All external routes are included in the topology table, and are tagged with the
More about this table later! following information:
RouterB#show ip eigrp topology • The identification number (router ID) of the EIGRP router that redistributed the
route into the EIGRP network
IP-EIGRP Topology Table for process 44
– The EIGRP router ID is normally selected in the same manner as Open
Codes: P - Passive, A - Active, U - Update, Q - Query, R -
Shortest Path First (OSPF)
Reply, r - Reply status
P 206.202.17.0/24, 1 successors, FD is 2195456
– Can also use: eigrp router-id <router-id>
via 206.202.16.1 (2195456/2169856), Ethernet0 • The AS number of the destination
P 206.202.18.0/24, 2 successors, FD is 2198016 • The protocol used in that external network
via 192.168.0.2 (2198016/284160), Serial0 • The cost or metric received from that external protocol
via 206.202.16.1 (2198016/2172416), Ethernet0 • The configurable administrator tag

6
 10/23/2012

Topology Table – Explained Soon! Topology Table

RTX#sh ip eigrp top 204.100.50.0

IP-EIGRP topology entry for 204.100.50.0/24 Much more on about this
State is Passive, Query origin flag is 1, 1 Successor(s), FD table and scenario later
is 2297856 after we discuss a few
Routing Descriptor Blocks: FD/RD more terms.
10.1.0.1 (Serial0), from 10.1.0.1, Send flag is 0x0
Composite metric is (2297856/128256), Route is External
Vector metric: SanJose2#show ip eigrp topology all-links
Minimum bandwidth is 1544 Kbit P 192.168.72.0/24, 1 successors, FD is 2172416, serno 93
Total delay is 25000 microseconds via 192.168.64.6 (2172416/28160), Serial0
Reliability is 255/255 via 192.168.1.1 (2174976/2172416), FastEthernet0/0
Load is 1/255 P 192.168.64.0/30, 1 successors, FD is 2172416, serno 91
Minimum MTU is 1500 via 192.168.1.1 (2172416/2169856), FastEthernet0/0
Hop count is 1 via 192.168.64.6 (2681856/2169856), Serial0
External data: P 192.168.64.4/30, 1 successors, FD is 2169856, serno 72
Originating router is 192.168.1.1 via Connected, Serial0
AS number of route is 0 P 192.168.1.0/24, 1 successors, FD is 28160, serno 1
External protocol is Connected, external metric is 0 via Connected, FastEthernet0/0
Administrator tag is 0 (0x00000000)

Topology Table IP Routing Table

Question: Since EIGRP has a topology table, does this make it a link- • EIGRP chooses the best routes (that is, successor) to a destination from the
state routing protocol? topology table and places these routes in the routing table.
• Each EIGRP router maintains a topology table for each network protocol.

Answer:
• EIGRP displays both internal EIGRP routes and external EIGRP routes.

• No, the information in the topology table is not in the form of LSAs RouterB#show ip route
describing the complete network topology. Codes: C - connected, S - static, I - IGRP, R - RIP, M - mobile, B - BGP
• The EIGRP topology table contains information about paths through D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
the router’s adjacent neighbors. E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP
• Also, EIGRP does not perform shortest-path calculation by calculating i - IS-IS, L1 - IS-IS level-1, L2 - IS-IS level-2, * - candidate
default U - per-user static route
the shortest-path tree, but instead uses the DUAL algorithm. Gateway of last resort is not set
C 10.1.1.0 is directly connected, Serial0
Alex Zinin, Cisco IP Routing D 172.16.0.0 [90/2681856] via 10.1.1.0, Serial0
D EX 192.168.1.0 [170/2681856] via 10.1.1.1, 00:00:04, Serial0

7
 10/23/2012

IP Routing Table Showing the cost in the Routing Table

• The routing table contains the routes installed by DUAL as the best loop-free
paths to a given destination.
• EIGRP will maintain up to four routes per destination.
• These routes can be of equal, or unequal cost (if using the variance
command). (later)
RouterB#show ip route
Codes: C - connected, S - static, I - IGRP, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP
i - IS-IS, L1 - IS-IS level-1, L2 - IS-IS level-2, * - candidate
default U - per-user static route
Gateway of last resort is not set
C 10.1.1.0 is directly connected, Serial0
D 172.16.0.0 [90/2681856] via 10.1.1.0, Serial0
D EX 192.168.1.0 [170/2681856] via 10.1.1.1, 00:00:04, Serial0 Routing Table
SanJose2#show ip route
D 192.168.72.0/24 [90/2172416]
via 192.168.64.6, 00:28:26, Serial0

Routing Tables EIGRP Technologies

RTX#show ip route
Gateway of last resort is 0.0.0.0 to network 0.0.0.0 Four key technologies set EIGRP apart from IGRP
D 192.168.5.0/24 [90/3219456] via 10.2.0.2, 00:12:19, Serial1
D 192.168.1.0/24 [90/2195456] via 10.1.0.1, 00:12:19, Serial0
C 192.168.2.0/24 is directly connected, Ethernet0
D 192.168.3.0/24 [90/2195456] via 10.2.0.2, 00:12:19, Serial1

RTX#show ipx route

11 Total IPX routes. Up to 1 parallel paths and 16 hops allowed.
No default route known.
C 1000 (HDLC), Se0
E 3000 [2681856/0] via 2000.0000.0c76.080c, age 00:10:49, 1u, Se1
E 4000 [276864000/2] via 2000.0000.0c76.080c, age 00:10:41, 1u, Se1

RTX#show appletalk route

Codes: R - RTMP derived, E - EIGRP derived, C - connected, A - AURP
S - static P - proxy
6 routes in internet
The first zone listed for each entry is its default (primary) zone.
E Net 100-101 [1/G] via 1000.123, 1400 sec, Serial0, zone san fran
E Net 300-301 [1/G] via 2000.28, 3016 sec, Serial1, zone san jose

8
 10/23/2012

1-Establishing Adjacencies with Neighbors Hello Intervals and Default Hold Times
Extra
• EIGRP routers establish adjacencies with neighbor routers by using small
hello packets.
• Hellos are sent every 5 seconds by default (224.0.0.10)
• K values must be the same between neighbors.
• An EIGRP router assumes that, as long as it is receiving hello packets from
known neighbors, those neighbors (and their routes) remain viable.
• Hold time tells the router how long it should consider the neighbor alive if it
has not received any EIGRP packets (Hello, EIGRP updates, etc.)
• Hello Time The interval of Hello Packets
• Hold time is normally three times the configured Hello interval. • Hold Time The interval to wait without receiving anything from a
neighbor before considering the link unavailable.
• Both the Hello and Hold time intervals are configurable on a per interface
basis, and do not have to match neighbor.
• EIGRP routers exchange routing information the same way as other distance
vector routing protocols, but do not send periodic updates.(Triggered
Updates)
• EIGRP updates are only sent when a network is added or removed from the
topology database, when the successor for a given network changes, or
when the locally used metric is updated. (later)
• EIGRP, like any other distance-vector routing protocol uses split-horizon.

Establishing Adjacencies with Neighbors 2-Reliable Transport Protocol (RTP)

• EIGRP is protocol-independent; that is, it doesn’t rely on TCP/IP to

exchange routing information the way RIP, IGRP, and OSPF do.
• To stay independent of IP, EIGRP uses the transport-layer protocol to
guarantee delivery of routing information: RTP.
• RTP supports reliable and unreliable delivery
• RTP supports unicasting and multicasting
• Initial delivery of EIGRP messages are done using multicast packets, that is
data is sent to all neighbors on a segment, and every neighbor is expected to
acknowledge it with a unicast Hello packet.
By forming adjacencies, EIGRP routers do the following: • After adjacency has been formed and added to neighbor table, routers
• Dynamically learn of new routes that join their network exchange routing information which is stored in the topology table. (later)
• Identify routers that become either unreachable or inoperable • RTP is used for EIGRP queries, updates and replies
• Rediscover routers that had previously been unreachable • RTP is not used for EIGRP Hello’s and Ack’s

9
 10/23/2012

3-DUAL FSM FSM Example

• The centerpiece of EIGRP is DUAL(Diffusing Update Algorithm )

the EIGRP route-calculation engine.
– The full name of this technology is DUAL finite state
machine (FSM). HUNGRY
(START)
NO FOOD
– This engine contains all the logic used to calculate and FOR 5
HOURS FOOD IS
compare routes in an EIGRP network. INEDIBLE
GET FOOD

What is FSM?
• An FSM is an abstract machine, not a mechanical device with EAT MORE
FULL
NOT ENOUGH
FOOD EATING FOOD
moving parts.
• FSMs define a set of possible states something can go through,
what events causes those states, and what events result from
those states.
• Designers use FSMs to describe how a device, computer EAT ENOUGH
FOOD
program, or routing algorithm will react to a set of input events.

DUAL FSM – Explained in a moment… 4-Protocol-Dependent Modules (PDMs)

• DUAL selects alternate routes quickly by using the information in the
EIGRP tables. • EIGRP is modular
• If a link goes down, DUAL looks for a feasible successor in its neighbor • Different PDMs can be added to EIGRP as new routed protocols are
and topology tables. enhanced or developed:
• A successor is a neighboring router that is currently being used for – IPv4, IPv6, IPX, and AppleTalk
packet forwarding, provides the least-cost route to the destination, and
is not part of a routing loop.
Each PDM is responsible for all functions related to its specific routed
• Feasible successors provide the next lowest-cost path without
protocol.
introducing routing loops.
– Feasible successor routes can be used in case the existing route • The IP-EIGRP module is responsible for the following:
fails; packets to the destination network are immediately forwarded – Sending and receiving EIGRP packets that bear IP data
to the feasible successor, which at that point, is promoted to the – Notifying DUAL of new IP routing information that is received
status of successor.
– Maintaining the results of DUAL’s routing decisions in the IP
• Selects a best loop-free path to a destination, the next hop being
routing table
known as the successor.
• All other routers to the same destination, that also meet the feasible – Redistributing routing information that was learned by other IP-
condition, meaning they are also loop-free (later), become feasible capable routing protocols
successors, or back-up routes.
• debug eigrp fsm

10
 10/23/2012

Protocol-Dependent Modules (PDMs) EIGRP Terminology and Operations

• EIGRP routers keep route and topology information readily

available in RAM so that they can react quickly to changes.

• Like OSPF, EIGRP keeps this information in several tables, or

databases.

– Neighbor table
– Topology table
– Routing table
– Successor
– Feasible Successor

• We will first have an overview of all of the terminology and then see
how it works and what it all means!

EIGRP Terminology and Operations Successors and Feasible Successors

• Successor – Current Route
– A successor is a route selected as the primary route to
use to reach a destination.
– Successors are the entries kept in the routing table.

• Feasible Successor - A backup route

– A feasible successor is a backup route.
– These routes are selected at the same time the
successors are identified, but they are kept in the
topology table.
– Multiple feasible successors for a destination can be
retained in the topology table.

Let’s see how this works!

11
 10/23/2012

Successors and Feasible Successors Successors and Feasible Successors

Successors and Feasible Successors Successors and Feasible Successors

Feasible distance (FD) is the minimum distance (metric) along a path to a
destination network.
172.30.1.0
Reported distance (RD) is the distance (metric) towards a destination as advertised
by an upstream neighbor. Reported distance is the distance reported in the
queries, the replies and the updates.
A neighbor meets the feasible condition(FC) if the reported distance by the
neighbor is smaller than or equal to the current feasible distance (FD) of this
router. "If a neighbors metric is less than mine, then I know the neighbor doesn't
have a loop going through me."
A feasible successor is a neighbor whose reported distance (RD) is less than the
current feasible distance (FD). Feasible successor is one who meets the feasible
condition (FC).

Your route (metric) to the network (RD to me) must be LESS than my current route
(my total metric) to that same network. If your route (metric) to the network (RD to 172.30.1.0
me) is LESS than my current route (my total metric), I will include you as a
FEASIBLE SUCCESSOR.
If your route (metric) to the network (RD to me) is MORE than my current route (my
total metric), I will NOT include you as a FEASIBLE SUCCESSOR.

12
 10/23/2012

Successors and Feasible Successors Successors and Feasible Successors

Feasible Successor, FC: RD30 < FD31 Feasible Successor, FC: RD30 < FD31

172.30.1.0 172.30.1.0

FD to 172.30.1.0 is FD to 172.30.1.0 is
31 via Router Y 31 via Router Y

Current Successor = 31
RTZ is NOT Feasible Current Successor = 31
RTZ is NOT Feasible
RD of RTY= 21 Successor, FC: RD of RTY= 21 Successor, FC:
RD220 not< FD31 RD220 not< FD31
Advertised or
Destination Feasible Dist. Reported. Dist. Neighbor • RTY is successor with a computed cost of 31.
172.30.1.0 40 30 X In Topology Table • “31” is the Feasible Distance (FD).
172.30.1.0 31 21 Y In Routing Table
• RTX is a feasible successor because its RD is less than or equal to the FD.
- RTX’s RD (30) is less than the FD (31).
172.30.1.0 230 220 Z Not in Topology Table

Example of a Loop, What if… What if the successor fails?

Feasible Successor exists:

RTZ FD = 220 172.30.1.0 • If current successor route fails, feasible successor becomes the
RTZ to RTA is current successor, i.e. the current route.
189 • Routing of packets continue with little delay.
RTA to
172.30.1.0 is No Feasible Successor exists:
31
• This may be because the Reported Distance is greater than the
Feasible Distance.
• Before this route can be installed, it must be placed in the active
RTZ has a Reported Distance to state and recomputed using DUAL FSM Algorithm. (later)
RTA of 220. Since its Reported
Distance is greater than RTA’s • Routing of packets continue but with more of a delay.
Cost=40 Cost=40
own Feasibile Distance of 31,
RTA can’t trust that the route
Cost=9 RTZ takes is somehow back
through itself.

13
 10/23/2012

Successors and Feasible Successors Successors and Feasible Successors

New Successor

172.30.1.0 X 172.30.1.0

FD to 172.30.1.0 is
X FD to 172.30.1.0 is
X
40 via Router X 40 via Router X

?
RTZ is NOT Feasible RTZ is NOT Feasible
Current Successor = 40 Current Successor = 40
RD of RTX= 30 Successor, FC: RD of RTX= 30 Successor, FC:
RD220 not< FD31 RD220 not< FD40
• Since RTX is the feasible successor, and becomes the successor. • RTZ is not a feasible successor.
• RTX is immediately installed from the topology table into the routing table (no recomputation • It’s RD (220) is greater than the previous FD (40) for 172.30.1.0/24.
of DUAL).
• RTA’s new FD via RTX is 40.
• Before this route can be installed, the route to net 24 must be placed in the
active state and recomputed using DUAL.
• RTZ is not a feasible successor, because it’s RD (220) is still greater than the new FD (40)
for 172.30.1.0/24. • Coming soon!

Successors and Feasible Successors One last reminder….

Topology table
X 172.30.1.0 • Each EIGRP router maintains a topology table for each configured
network protocol.
• This table includes route entries for all destinations that the router has
learned. All learned routes to a destination are maintained in the
topology table.

FD to 172.30.1.0 is
X
230 via Router Z
show ip eigrp topology
• (Feasible Distance/Reported Distance)
RTZ is NOT Feasible
Current Successor = 230
RD of RTZ= 220 Successor, FC: • 1 successor (route) if FDs are different
RD220 not< FD40 – smaller FD metric, that route is the the only successor
– larger FD metric, those routes are possible feasible successor
• After a a series of EIGRP Queries and Replies (coming), and a recomputation • 2 or more successors (routes) if FDs are the same
of DUAL, RTZ becomes the successor. – Load balancing happens automatically
• There is nothing better to prohibit it from being the successor.

14
 10/23/2012

SanJose2#show ip eigrp top all Understanding the

P 192.168.72.0/24, 1 successors, FD is 2172416, serno 93
via 192.168.64.6 (2172416/28160), Serial0 Topology Table
via 192.168.1.1 (2174976/2172416), FastEthernet0
P 192.168.64.0/30, 1 successors, FD is 2172416, serno 91
via 192.168.1.1 (2172416/2169856), FastEthernet0
via 192.168.64.6 (2681856/2169856), Serial0

SanJose2#show ip eigrp top all

P 192.168.72.0/24, 1 successors, FD is 2172416, serno 93
Understanding the Topology Table via 192.168.64.6 (2172416/28160), Serial0
via 192.168.1.1 (2174976/2172416), FastEthernet0
(not in the curriculum) P 192.168.64.0/30, 1 successors, FD is 2172416, serno 91
via 192.168.1.1 (2172416/2169856), FastEthernet0
via 192.168.64.6 (2681856/2169856), Serial0
P 192.168.64.4/30, 1 successors, FD is 2169856, serno 72
via Connected, Serial0
P 192.168.1.0/24, 1 successors, FD is 28160, serno 1
via Connected, FastEthernet0

Understanding the SanJose2#show ip eigrp top all-links

IP-EIGRP Topology Table for AS(100)/ID(192.168.64.5)
Topology Table Codes: P - Passive, A - Active, U - Update, Q - Query, R
- Reply,r - Reply status
Feasible Distance/Reported Distance Feasible distance
P 192.168.72.0/24, 1 successors, FD is 2172416, serno 22
successor via 192.168.64.6 (2172416/28160), Serial0
feasible via 192.168.1.1 (2174976/2172416), FastEthernet0
SanJose2#show ip eigrp top all-links
successor
IP-EIGRP Topology Table for AS(100)/ID(192.168.64.5)
Codes: P - Passive, A - Active, U - Update, Q - Query, R Reported Distance
- Reply,r - Reply status
Feasible distance Reported Distance: This is the distance (cost) reported by the neighboring router, ie.
Westasman and SanJose1.
P 192.168.72.0/24, 1 successors, FD is 2172416, serno 22
• Westasman’s Reported Distance is 28,160 = 25,600 (BW) + 2,560 (DLY)
successor via 192.168.64.6 (2172416/28160), Serial0
feasible
• SanJose1’s Reported Distance is 2,172,416 = 1,657,856 (BW) + 512,000 (DLY) +
via 192.168.1.1 (2174976/2172416), FastEthernet0 2,560 (DLY)
successor
Feasible Successor: Since SJ1’s Reported Distance 2172416 is less than or
Feasible distance if this router was the successor. Includes the 2,560
equal to the Feasible Distance 2172416 , it is included as a Feasible
delay cost of additional Fast Ethernet link.
Successor!

15
 10/23/2012

Questions Example of a Loop, What if…

Question: What if there is only one entry in the Topology Table?

RTZ FD = 220 172.30.1.0
Answer: There is only a successor and no feasible successors.
• There are either no other neighbors with a route to this network or the RTZ to RTA is
reported distance of that neighbor is greater than the current feasible 189
distance. (You will see this in the lab.) RTA to
172.30.1.0 is
31
Question: Why does EIGRP use DUAL? I.e. Why doesn’t EIGRP install
routes with an RD greater than its current FD?
Answer: EIGRP is a distance vector routing protocol. It only knows about
distances/routes from what its’ neighbors tell it. The only way EIGRP can
RTZ has a Reported Distance to
trust that another router’s route is not back through itself, is to make sure RTA of 220. Since its Reported
the other router’s distance is equal to or less than its own distance to that Distance is greater than RTA’s
network. Cost=40 Cost=40
own Feasibile Distance of 31,
RTA can’t trust that the route
Cost=9 RTZ takes is somehow back
through itself.

EIGRP Packet Types EIGRP Hello Packet

The five EIGRP packet types are: • Used to discover, verify, and rediscover neighbor routers.
• EIGRP routers send hellos at a fixed (and configurable) interval, called
the hello interval.
• The default hello interval depends on the bandwidth of the interface.
• Hello interval 5 seconds, hold time 15 seconds for T1 and faster
• Hello interval 60 seconds, holdtime 180 seconds(3min)for slower than T1
• On IP networks, EIGRP hello packets are multicast, 224.0.0.10
• If a neighbor is not heard from for the duration of the hold time (three
times hello interval), EIGRP considers that neighbor down, and DUAL
must step in to reevaluate the routing table.
–By default, the hold time is three times the hello interval, but an
administrator can configure both timers as desired.
• Unlike OSPF routers, EIGRP (Neighbor) routers do not need to have the
same hello intervals and hold down intervals.

16
 10/23/2012

Acknowledgement Packet Update Packet

• Acknowledgement packets, which are “data-less” hello packets, are • Update packets are used when a router discovers a new neighbor.
used to ensure reliable communication. –An EIGRP sends unicast update packets to that new neighbor so
• Unlike multicast hellos, acknowledgement packets are unicast. that it can add to its topology table.
• Acknowledgements can be made by piggybacking on other kinds of –More than one update packet may be needed to convey all of the
EIGRP packets, such as reply packets. topology information to the newly discovered neighbor.
• EIGRP updates are only sent when:
-A network is added or removed from the topology database
-The successor for a given network changes
-The locally used metric is updated.
• The EIGRP router sends a multicast update packet (224.0.0.9) to all
neighbors alerting them to the change.
• EIGRP routers exchange routing information the same way as other
distance vector routing protocols, but do not send periodic updates.

Query and Reply Packets Query and Reply Packets

• A router views its feasible successors as neighbors that are downstream, or
• EIGRP routers use query packets whenever it needs specific information closer, to the destination than it is.
from one, or all, of its neighbors. • If something goes wrong with the successor, DUAL can quickly identify a
– A reply packet is used to respond to a query. feasible successor from the topology table, and install a new route to the
destination.
• If an EIGRP router loses its successor and cannot find a feasible
successor for a route, DUAL places the route in the active state. • If no feasible successors to the destination exist, DUAL places the route in the
active state.
– The router multicasts a query to all neighbors, searching for a
successor to the destination network.
• Entries in the topology table can be in one of two states: active or passive.
– Neighbors must send replies that either provide information on • A passive route is one that is stable and available for use.
successors, or indicate that no successor information is available. • An active route is a route in the process of being recomputed by DUAL.
• Queries can be multicast or unicast, while replies are always unicast.
• Recomputation happens if a route becomes unavailable and DUAL can’t find
any feasible successors.
• Another route may exist, it is just that their Reported Distance was greater than
your Feasible Distance.

17
 10/23/2012

Return Route or Forward Query

Query and Reply Packets RtrD
If a feasible successor does not exist:
• 1. The router flags the route as active.
RtrB
• 2. The router looks for an alternate path by sending out a query packet to all neighbors to
Queries learn if they have a path to the given destination.
Replies – The query packets are multicast out every interface except the one which the dead link
RtrE was learned, adhering to the split horizon rule.
RtrA
• 3. If a neighbor does have a path that does not involve the querying router, or no path at all
to the destination, it unicasts a reply with this information.
X – If a neighbor that receives the query is using the querying router as its feasible
successor, then it multicasts its own query packet to its neighbor, which creates a
RtrF “ripple effect” through the network until a new path is found or a major network
boundary is met.
RtrC • 4. When the query router receives replies, it reacts based on the answer in the reply:
– If the reply included a successor or feasible successor, the information is put into its
RtrG topology table, and the querying router waits until all replies are received. It then
recalculates the topology table, and adds the successr(s) to the routing table. The route
returns to a passive state in the topolgy table and routing can continue.
Looking for new route – If none of the replies includes a successor or feasible successor, the querying router
• If there were no Feasible Successors, the router must ask neighbors for help in hope of removes the active route from its topology table and routing tables.
finding a new, loop-free path to the destination. • If a neighbor router to which a query is sent does not reply within the active time of 180
• Neighbor routers are compelled to reply to this query. seconds(3min), EIGRP tears down the neighbor relationship with the offending router and
– If a neighbor has a route, it will reply with information about the successor(s). puts routes learned from that router into an active state.
– If not, the neighbor notifies the sender that it doesn’t have a route to the
destination either.

Query and Reply Packets In this scenario…

X 172.30.1.0 X
Queries Queries
Replies Replies

Routes via RTY

X All
X
and RTX Fail! Replies

? are
saying
they do
?
RTZ was previously not have
NOT a Feasible a route
Successor, FC: RD220 RTZ has a Reported Distance to
not< FD31 or FD40, but RTA of 220. Since its Reported
now there is no Distance is greater than RTA’s
Cost=89 Cost=99
Sucessor own Feasibile Distance of 31,
RTZ replies that it still has a route to 172.30.1.0, while RTX and RTY reply RTA can’t trust that the route
that they do not. Cost=100 RTZ takes is somehow back
through itself.
Current Successor is now RTZ, with a FD of 230 and a RD of RTZ= 220.

18
 10/23/2012

Example from the curriculum… Example from the curriculum…

1 2 5 6

3 4
7

Example of debug eigrp fsm Stuck in Active (SIA)

Router
C
Router
X D
Router
No feasible successor in the topology table. EIGRP domain still finds another route.
SanJose2#debug eigrp fsm
EIGRP FSM Events/Actions debugging is on
SanJose2(config)#inter s 0
Queries
Replies
B
X
SanJose2(config-if)#shut
Router
A X
03:11:44: DUAL: Destination 192.168.72.0/24
03:11:44: DUAL: Find FS for dest 192.168.72.0/24. FD is 2172416, RD is X
2172416
03:11:44: DUAL: 192.168.64.6 metric 4294967295/4294967295 not
found Dmin is 4294967295
03:11:44: DUAL: Dest 192.168.72.0/24 entering active state.
X
Feasible successor is in the topology table. Backup route takes over right away.
Westasman#debug eigrp fsm
02:21:42: DUAL: Find FS for dest 192.168.64.4/30. FD is 2169856, RD is
 In some cases, it can take too long for the query to be answered.
2169856
02:21:42: DUAL: 0.0.0.0 metric 2169856/0  When this happens, the router that issued the query gives up and resets its
02:21:42: DUAL: 192.168.64.1 metric 4294967295/4294967295 neighbor relationship with the router that didn’t answer.
found Dmin is 216985  The most basic situation where this occurs is when it simply takes too long
for a query to reach the other end of the network and a reply to travel back.

19
 10/23/2012

Stuck in Active (SIA) Troubleshooting (SIA) - FYI

• Typically, SIAs results when a router cannot answer a Troubleshooting Steps:

Step 1: find the routes which are consistently being reported as stuck in
query because: active.
– If you are logging console messages, a quick perusal of the log will
– the router is too busy to answer the query (generally indicate which routes are being marked as stuck in active most
high cpu utilization) often.
Step 2: find out which routers are consistently failing to answer queries (not
– the router cannot allocate the memory to process the always easy).
query or build the reply packet – Use the show ip eigrp topology active command.
– the circuit between the two routers is not good (packet • Any neighbors which have the r beside them are neighbors that
the router is waiting on replies from
loss) • the active timer is how long the route has been active.
– unidirectional links (a link on which traffic can only flow • pay particular attention to routes that have replies outstanding
and have been active for 2 to 3 minute
in one direction due to a failure) Step 3: find the reason why that router is not receiving or answering queries
– One you have found the router that is consistently not answering
queries, look for problems on the link to this neighbor, memory or
CPU utilization problems with this neighbor, etc.

Configuring EIGRP for IP networks

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

• This value must match all routers within the internetwork.

Router(config-router)#network network-number
• The network command configures only connected networks.

Router(config-router)#eigrp log-neighbor-changes
• This command enables the logging of neighbor adjacency changes to
monitor the stability of the routing system and to help detect problems.

Router(config-if)#bandwidth kilobits
• When configuring serial links using EIGRP it is important to configure
Configuring EIGRP the bandwidth setting on the interface. If the bandwidth setting is not
changed for these interfaces EIGRP assumes the default bandwidth
on the link instead of the true bandwidth.

20
 10/23/2012

Summarizing EIGRP Routes: Summarizing EIGRP Routes:

no auto-summary no auto-summary

• EIGRP automatically summarizes routes at the classful

boundary, the boundary where the network address ends
as defined by class-based addressing.
• In the presence of discontiguous subnetworks, automatic
summarization must be disabled for routing to work properly.
• To turn off auto-summarization, use the following command:
Router(config-router)#no auto-summary

Summarizing EIGRP Routes: Interface Summarizing EIGRP Routes: Interface

Summarization Summarization

Router(config-if)#ip summary-address eigrp autonomous-system-number RTC(config)#interface serial0/0

ip-address mask administrative-distance RTC(config-if)#ip summary-address eigrp 2446 2.1.0.0 255.255.0.0

RTC(config)#router eigrp 2446 RTC’s Routing Table:

D 2.1.0.0/16 is a summary, 00:00:22, Null0
RTC(config-router)#no auto-summary
RTC(config-router)#exit
• Notice that the summary route is sourced from Null0, and not an actual interface.
• That is because this route is used for advertisement purposes and does not represent a
RTC(config)#interface serial0/0 path that RTC can take to reach that network.
RTC(config-if)#ip summary-address eigrp 2446 2.1.0.0 255.255.0.0 • On RTC, this route has an administrative distance of 5.
• RTD is oblivious to the summarization but accepts the route. It assigns the route the
administrative distance of a "normal" EIGRP route, which is 90, by default

21
 10/23/2012

EIGRP show commands EIGRP debug commands

EIGRP variance command

Optional Interface Command:

RTA(config-router)# variance number

IGRP and EIGRP also support unequal cost path load

balancing, which is known as variance.
The variance command instructs the router to include routes
Few last items not in the curriculum… with a metric less than or equal to n times the minimum metric
route for that destination, where n is the number specified by
the variance command.
Note: If a path isn't a feasible successor, then it isn't used in
load balancing.

22
 10/23/2012

EIGRP variance command EIGRP variance command

router eigrp 1
network x.x.x.x
variance 2
variance 2
Let's look at an example, there are three ways to get to Network X, from
Router E: (Note: metrics incorrectly not including outgoing interface to • This increases the minimum metric to 40 (2 * 20 = 40).
Net X.) • EIGRP includes all the routes that have a metric less than or equal
 E-B-A with a metric of 30 to 40, and are feasible successors.
 E-C-A with a metric of 20 <<<-- Minimum Metric • In the above configuration, EIGRP now uses two paths to get to
Network X, E-C-A and E-B-A, because both paths have a metric
 E-D-A with a metric of 45 under 40.

EIGRP variance command EIGRP Traffic Share

• To control how traffic is distributed among routes when there are
multiple routes for the same destination network that have different
costs, use the traffic-share router configuration command.
• To disable this function, use the no form of the command.

traffic-share {balanced | min}

router eigrp 1 no traffic share {balanced | min}
network x.x.x.x
variance 2 balanced
• Distributes traffic proportionately to the ratios of the metrics.
Two unequal-cost paths: min
 E-B-A with a metric of 30 <<<-- Also below 2*20 • Uses routes that have minimum costs
 E-C-A with a metric of 20 <<<-- Minimum Metric
 E-D-A with a metric of 45 • For more information:
http://www.cisco.com/en/US/tech/tk365/tk207/technologies_tech_note0
9186a008009437d.shtml
EIGRP doesn't use path E-D-A because it has a metric of 45, and it's not a feasible
successor. Can have up to 6 unequal cost paths.

23
 10/23/2012

Redistribution b/t EIGRP & IGRP EIGRP and Default Routes (Review)

There are three ways to inject a default route into

EIGRP:
• Redistribute a static route
• IP default-network
Same AS numbers Different AS numbers
Router Two Router Two
• Summarize to 0.0.0.0/0
router eigrp 2000 router eigrp 2000
network 172.16.1.0 redistribute igrp 1000
! network 172.16.1.0
router igrp 2000 !
network 10.0.0.0 router igrp 1000
redistribute eigrp 2000
(automatic redistribution) network 10.0.0.0

So what is the difference? We’ll see later when discussing Redistribution.

EIGRP and Default Routes (Review) EIGRP and Default Routes (Review)

Redistribute a static route Ip default-network

• Use the first method when you want to draw all traffic to • Propagates a default route to other routers, but needs to have a
unknown destinations to a default route at the core of the route or default route out once the packets arrive.
network.
• This method is effective for advertising connections to the Gateway Router
Internet, but will redistribute all static routes into EIGRP.
router igrp 24
• For example: <text omitted>
network 207.21.20.0
Gateway Router
ip route 0.0.0.0 0.0.0.0 x.x.x.x (next hop) ip route 0.0.0.0 0.0.0.0 207.21.20.1
! ip default-network 207.21.20.0
router eigrp 100
redistribute static
<text omitted>

24
 10/23/2012

EIGRP and Default Routes (Review) OSPF versus EIGRP

OSPF EIGRP

Supports CIDR and VLSM, rapid Supports CIDR and VLSM, rapid
Extra: Summarize to 0.0.0.0/0 convergence, partial updates, neighbor convergence, partial updates, neighbor
• Summarizing to a default route is effective only when you want discovery discovery
to provide remote sites with a default route. Administrator can define route Automatic route-summarization and user-
summarization defined route summaries
• Since summaries are configured per interface, you don't need to
worry about using distribute-lists or other mechanisms to Open standard; multivendor support Proprietary; Cisco routers only
prevent the default route from being propagated toward the core
of your network.
Scalable; administratively defined Scalable, but no hierarchical design
“areas” provide manageable hierarchy
router eigrp 100
network 10.0.0.0
!
Difficult to implement Easy to implement
interface serial 0
ip address 10.1.1.1
ip summary-address eigrp 100 0.0.0.0 0.0.0.0
Equal-cost load balancing Unequal-cost load balancing

Metric Calculation
The metrics used by EIGRP in making routing decisions are (lower the metric the
better):
• bandwidth
• delay
• load
• reliability

By default, EIGRP uses only:

EIGRP and IGRP Metric Calculation • Bandwidth
• Delay
Extra Material
Analogies:
Think of bandwidth as the width of the pipe
and
delay as the length of the pipe.

• Bandwidth is a the carrying capacity

• Delay is the end-to-end travel time.

25
 10/23/2012

Metric Calculation Metric Calculation

If these are the default:
• bandwidth (default) Router> show interfaces s1/0
• delay (default) Serial1/0 is up, line protocol is up
Hardware is QUICC Serial bandwidth delay
When are these used? Description: Out to VERIO
• load Internet address is 207.21.113.186/30
• reliability MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
rely 255/255, load 246/255
Only when configured by the network administrator to do so!
Encapsulation PPP, loopback not set
EIGRP also tracks (but does not use in its metric calculation):
Keepalive set (10 sec)
• MTU (Maximum Transmission Unit)
<output omitted>
• Hop Count
reliability load

Use show interface command to view the metrics used on a specific interface
that is routing EIGRP.
• These are the raw values.

Metric Calculation Metric Calculation

Bandwidth
Changing the bandwidth informational parameter:
• Expressed in kilobits (show interface)
• This is a static number and used for metric calculations only.
The bandwidth can be changed using:
• Does not necessarily reflect the actual bandwidth of the link. Router(config-if)# bandwidth kilobits
• It is an information parameter only.
• You cannot adjust the actual bandwidth on an interface with this command.
To restore the default value:
• Use the show interface command to display the raw value Router(config-if)# no bandwidth

The default values:

• Default bandwidth of a Cisco interface depends on the type of interface.

• Default bandwidth of a Cisco serial interface is 1544 kilobits or 1,544,000

bps (T1), whether that interface is attached to a T1 line (1.544 Mbps) or a 56K
line.

• IGRP/EIGRP metric uses the slowest bandwidth of all of the outbound

interfaces to the destination network.

26
 10/23/2012

Metric Calculation Metric Calculation

Delay Changing the delay informational parameter:
• Like bandwidth, delay it is a static number. The delay can be changed using:
• Expressed in microseconds, millionths of a second Router(config-if)# delay tens-of- S
• (Uses the Greek letter mu with an S, S, NOT “ms” which is millisecond or (microseconds)
thousandths of a second)
• Use the show interface command to display the raw value
• It is an information parameter only. Example of changing the delay on a serial interface to 30,000
microseconds:
The default values: Router(config-if)# delay 3000
• The default delay value of a Cisco interface depends upon the type of
interface. To restore the 20,000 microsecond default value:
Router(config-if)# no delay
• Default delay of a Cisco serial interface is 20,000 microseconds, that of a T1
line.

• IGRP/EIGRP metric uses the sum of all of the delays of all of the outbound
interfaces to the destination network.

Metric Calculation Metric Calculation

IGRP
• bandwidth = (10,000,000/bandwidth) Router> show interfaces s1/0
• delay = delay/10 Serial1/0 is up, line protocol is up
Hardware is QUICC Serial bandwidth delay
EIGRP Description: Out to VERIO
• bandwidth = (10,000,000/bandwidth) * 256 Internet address is 207.21.113.186/30
• delay = (delay/10) * 256
MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
Note: The reference-bandwidth rely 255/255, load 246/255
For both IGRP and EIGRP: 107, (10,000,000/bandwidth kbps), whereas Encapsulation PPP, loopback not set
with OSPF it was 108 (100,000,000/bandwidth)
Keepalive set (10 sec)
<output omitted>
The difference:
• IGRP metric is 24 bits long reliability load
• EIGRP metric is 32 bits long
• EIGRP metric is 256 times greater for the same route
• EIGRP allows for finer comparison of potential routes

27
 10/23/2012

Reliability and Load Reliability and Load

The metrics used by EIGRP in making routing decisions are (lower the Reliability
metric the better): • Reliability is measure dynamically
• bandwidth • Uses error rate for measurement
• delay • Reflects the total outgoing error rates of the interfaces along the
• load route
• reliability • Calculated on a five minute weighted average, so not to allow
sudden peaks and valleys to make a significant impact
By default, EIGRP uses only:
• Bandwidth Expressed as an 8 bit number
• Delay • 255 is a 100% reliable link
• 1 is a minimally reliable link

Higher the better!

Reliability and Load Reliability and Load

Router> show interfaces s1/0 Load
Serial1/0 is up, line protocol is up • Load is measure dynamically
Hardware is QUICC Serial bandwidth delay • Uses channel occupancy for measurement
Description: Out to VERIO • Reflects the total outgoing load of the interfaces along the route
Internet address is 207.21.113.186/30 • Calculated on a five minute weighted average, so not to allow sudden
MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec, peaks and valleys to make a significant impact
rely 255/255, load 246/255 Expressed as an 8 bit number
Encapsulation PPP, loopback not set • 255 is a 100% loaded link
Keepalive set (10 sec) • 1 is a minimally loaded link
<output omitted>

reliability load Lower the better!

shows reliability as a fraction of 255, for example: Note: Even though load and reliability are dynamically changing values,
rely 190/255 (or 74% reliability) EIGRP will not recalculate the route metric when these parameters
rely 234/255 (or 92% reliability)
change.
rely 255/255 (or 100% reliability)

28
 10/23/2012

Reliability and Load Reliability and Load

Router> show interfaces s1/0 IGRP/EIGRP metric =

[k1* BW IGRP(minimum) +
Serial1/0 is up, line protocol is up
bandwidth delay (k2* BW IGRP(minimum))/(256-LOAD) +
Hardware is QUICC Serial
k3* DLYIGRP(sum) ] *
Description: Out to VERIO [k5/RELIABILITY + k4)]
Internet address is 207.21.113.186/30 • k2 metric effects LOAD
MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec, • k4 and k5 effects RELIABILITY
rely 255/255, load 246/255 • Multiply Reliability only if > 0
Encapsulation PPP, loopback not set
Keepalive set (10 sec) Default:
k1=k3=1 and k2=k4=k5=0
<output omitted>
• You may change the k values to change what you want to give more or less weight to.
reliability load – k1 for bandwidth
shows load as a fraction of 255, for example: – k2 for load
load 10/255 (or 3% loaded link) – k3 for delay
load 40/255 (or 16% loaded link) – k4 and k5 for Reliability
load 255/255 (or 100% loaded link) • Higher the k value, the more that part of the metric is used to calculate the overall IGRP
metric

Reliability and Load

Turning the knobs:
We can use the other metrics of Reliability and Load by adjusting their k values
to something greater than “0”

The command to adjust the k values is:

Router(config-router)# metric weights tos k1 k2 k3 k4 k5

Notes:
• tos is always set to 0; at one time it was Cisco’s intent to use it, but it was
never implemented
• EIGRP neighbors must agree on K values to establish an adjacency and to
avoid routing loops.

Caution!
• Know what the impact will be before changing the defaults.
• It can give you unexpected results if you do not know what you are doing!
• If you modify the weights, you should configure all routers so they are all using
the same weight values.

29