Routing Algorithms

Network Layer 4-2

Dynamic Routing

The routing table is updated using the routing protocols. When there is a change in the Internet such as router breakdown or link failure, the routing protocols update all the information in the routing tables.

For the dynamic routing, routers should exchange routing information using routing protocols. routing information

• information about network topology, delay, etc. routing protocol

• It specifies the routing messages and procedures to exchange the routing information to determine routes.

Network Layer 4-3

Internet Routing Architecture Internet architecture from routing’s views

It is unrealistic to apply a single routing protocol to the worldwide Internet because of its size.

So, the worldwide Internet is divided into many groups, which are administered independently.

These independent groups of networks are called the autonomous systems(AS) which are assigned 16 bits long AS number.

AS • “a group of networks and routers controlled by a single

administrative authority” Each AS needs to inform its routing information of other

ASs. For this purpose each AS has more than one border routers.

Network Layer 4-4

Subnet 1.2

Subnet 1.4 Subnet

2.3

Subnet 2.2

Subnet 2.4

Subnet 2.1

Subnet 1.3

R1

R2R3

R4

R5

R6

R7

R8

Autonomous System 1 Autonomous System 2

LEGEND: Interior Gateway Protocol

Exterior Gateway Protocol

Subnet 1.1

Network Layer 4-5

Internet Routing Protocols

Interior Gateway Protocol (IGP) IGP is operated within each AS. Each AS can operate its own IGP. Most well-known IGPs

• RIP(Routing Information Protocol)• OSPF(Open Shortest Path Find)

Exterior Gateway Protocol (EGP) To exchange packets between AS, the AS border

routers should exchange the routing information. EGP is the routing protocol between ASs.

• BGP(Border Gateway Protocol)

Network Layer 4-6

IGPs

Routingprotocol

Shortest path algorithm

Routingalgorithm

RIPIGRP

Bellman-Ford algorithm

Distance Vector algorithm

OSPFIS-IS

Dijstra algorithm Link State algorithm

Network Layer 4-7

Distance Vector Algorithm

(assume that the metric is delay)Step 1: Each router exchange ECHO packet to measure the time to reach each neighboring routers.

Step 2: Each router send this information to its neighboring routers periodically. The information contains the times to reach all other routers.

Step 3: Each router determines the mim. time to reach all other routers using step 1 and 2. And update its own routing table.

Network Layer 4-8

Example

J A CEcho 패킷

m secTi sec

E

B

D(TAB, TAC, TAD, TAE)

TB = m + TABTC = m + TACTD = m + TADTE = m + TAE

Network Layer 4-9

Example: sample network

A

E

I

B C D

F G H

J K L

Network Layer 4-10

new information from A, I, H, K

From to

A I H K

A 0 24 20 21 B 12 36 31 28 C 25 18 19 36 D 40 27 8 24 E 14 7 30 22 F 23 20 19 40 G 18 31 6 31 H 17 20 0 19 I 21 0 14 22 J 9 11 7 10 K 24 22 22 0 L 29 33 9 9

Network Layer 4-11

New information on the neighboring nodes Suppose that the delay from J to A, I,H, and K

8,10,12,6 secs respectively. Then what is the next hop router and the delay to

reach G from J?

J ---> A -------> G 8 + 18 = 26

J ----> I -------> G 10 + 31 = 41

J ----> H ------> G 12 + 6 = 18 (best route for JG)

J ----> K ------> G 6 + 31 = 37

Network Layer 4-12

Updated routing table of J

destination delay Outgoing link A 8 A B 20 A C 28 I D 20 H E 17 I F 30 I G 18 H H 12 H I 10 I J 0 - K 6 K L 15 K

Network Layer 4-13

Link state Routing

Overview Each node collects the topology information of the

whole network, and then computes the shortest paths to reach each node using the Dijkstra algorithm.

The routing protocols such as ISO’s IS-IS and IETF’s OSPF belong to this algorithm, and they will replace the DVA routing protocol.

Examples OSPF for TCP/IP ISO’s IS-IS for CLNS and IP DEC’s DNA Phase V Novell’s NLSP(Netware Link services protocol)

Network Layer 4-14

Link state Routing Algorithm

Discover the neighboring node. (send Hello message periodically)

Measure the link costs to the neighboring nodes. Make the Link State packet. Broadcast the Link State packet to all nodes. After receiving the Link State packets from all other

nodes, make the link state database. Compute the shortest paths to reach all other

nodes based on the link state database.

Network Layer 4-15

Example of Link state routingStep1: Collect the link state information from the

neighboring nodes and make the link state packets.

A

B

D

C

E

F

5

2

3 1

2 1

1 332

Aseq#age

B 2C 5D 2

Dseq#age

A 2B 2C 3E 1

Cseq#age

A 5B 3D 3E 1F 1

Bseq#age

A 2C 3D 2

Eseq#age

C 1D 1F 1

Fseq#age

C 1E 1

Link state packets

Network Layer 4-16

Example of Link state routing

Step 2: Propagate the link state information to all other nodes. Make the link state packet. Use the flooding.

Step 3: Compute the shortest path. Based on the link state information, the node makes

the link state database that represents the whole network topology.

Compute the shortest path using the Dijkstra algorithm.

Network Layer 4-17

Link State Database

A

B

D

C

E

F

5

2

3 1

2 1

1 332

Link # Cost Link # CostA-BA-CA-DB-AB-CB-DC-A

2522325

D-EE-CE-DE-FE-CE-E

111111

C-BC-DC-EC-FD-AD-BD-C

3311223

Link # Cost

Network Layer 4-18

Internet Routing Protocols

Network Layer 4-19

Intra-AS Routing

Also known as Interior Gateway Protocols (IGP) Most common Intra-AS routing protocols:

RIP: Routing Information Protocol

OSPF: Open Shortest Path First

IGRP: Interior Gateway Routing Protocol (Cisco proprietary)

Network Layer 4-20

RIP IETF standard routing protocol Based on the Distant Vector algorithm The hop count is used as the metric.

The hop count is 1 for an adjacent network. The hop count 16 means infinity.(network is not

connected.) Each router exchange routing information every

30 secs. The routing information contains the whole routing table

entries. Each advertisement: list of up to 25 destination

nets within AS It uses the UDP port number 520. It was first implemented in the UNIX 4.2 BSD.

Network Layer 4-21

Procedure

Receive a response RIP message

1. Add one hop to the hop count for each advertised destination2. Repeat the following steps for each advertised destination:

If (destination not in the routing table)Add the advertised information to the table

elseif (next-hop field is the same)

replace entry in the table with the advertised oneelse

if (advertised hop count smaller than one in the tableadd it to the routing table

elsedo nothing

3. Return

Network Layer 4-22

RIP: Example(1)

Network Layer 4-23

RIP: Example(2)

Network Layer 4-24

RIP: Example(2)

Network Layer 4-25

RIP message

Command version

family

reserved

All 0s

Network address

All 0s

All 0s

distance

Command : request (1) or response (2) The unsolicited response is transmitted periodically every 30 secs. Family: TCP/IP (2)distance : hop count

반복

Network Layer 4-26

Problems of DVA(RIP)

Problems scalability Slow convergence

• In particular, when the shortest paths are changing rapidly, the inconsistency between routing tables can happen, since the updated routing information propagates slowly.

• count-to-infinity problem Not proper to be expanded to multicast

routing protocol.

Network Layer 4-27

Good news go travels quickly!

good news go fast! When the connection with the network 1 is added, the routing

table can reach the stability after 2 message exchanges.

upR1 R3R2Net 1

Net D R 1 1 R1

Net D R 1 -

Net D R 1 -

Net D R 1 1 R1

Net D R 1 2 R1

Net D R 1 -

Net D R 1 1 R1

Net D R 1 2 R1

Net D R 1 3 R2

(N1,1)

(N1,1)

(N1,2)(N1,2)

Initial state

time

Note: (x, y) = (destination network, Distance)

distance next routerdestination

Network Layer 4-28

Count-to-infinity Problem(1)crash

R1 R3R2Net 1

Net D R 1 -

Net D R 1 2 1

Net D R 1 3 2

Net D R 1 3 2

Net D R 1 4 3

Net D R 1 3 2

Net D R 1 5 2

Net D R 1 4 3

Net D R 1 5 2

Net D R 1 5 2

Net D R 1 6 1

Net D R 1 5 2

Net D R 1 -

Net D R 1 -

Net D R 1 -

(N1,2)

(N1, )

(N1,4)

(N1,3)

(N1,4)

(N1,5)

(N1,3)

(N1,2)

(N1,3)

(N1,4)

(N1,5)

(N1,4)... ... ...

Initial state

time

Network Layer 4-29

Count-to-infinity Problem(2) Bad news goes slowly!!

In the previous picture, when the connection between R1 and network 1 is broken, the cost (distance) to network 1 grows to infinity gradually.

When the cost reaches to 16, the router knows that the connection is broken.

When the bad news occur, it takes a long time to be aware of that.(slow convergence)

Solutions Split Horizon Hold down Poison reverse

Network Layer 4-30

Solutions to the Count-to-infinity Problem

Split horizon update (with Poison Reverse) The router interface that receives the information

about a certain network should send information about that network to other routers as infinity (distance=16)

It cannot be applied to all kinds of topologies. Route poisoning

If a router receives the information that the hop count to a certain network increases, it sets the hop count=16, and send it to other routers, suspecting a loop might occur.

Network Layer 4-31

Split Horizon

A router should never send the routing update information to the interface through which it received the corresponding routing information.

Net1 Net2 Net3

Net2 1Net3 2 Net3 1

A B

Net1 1Net1 2Net2 1

routing message routing message

Network Layer 4-32

Poison reverse

When a router sends the routing update message to all routers, it says to the interface from which it received the information about the network that the cost to the network is infinity (cost=16).

Net1 Net2 Net3

Net1 16Net2 1Net3 2

A B

routing message

Net1 1Net2 16Net3 16

Net1 16Net2 16Net3 1

routing message

Net1 2Net2 1Net3 16

Network Layer 4-33

RIP: Link Failure and Recovery If no advertisement heard after 180 sec -->

neighbor/link declared dead routes via neighbor invalidated new advertisements sent to neighbors neighbors in turn send out new advertisements

(if tables changed) link failure info quickly propagates to entire net poison reverse used to prevent ping-pong

loops (infinite distance = 16 hops)

Network Layer 4-34

RIP Table processing

RIP routing tables managed by application-level process called route-d (daemon)

advertisements sent in UDP packets, periodically repeated

physical

link

network forwarding (IP) table

Transprt (UDP)

routed

physical

link

network (IP)

Transprt (UDP)

routed

forwardingtable

Network Layer 4-35

OSPF (Open Shortest Path First)

“open”: publicly available Uses Link State algorithm

LS packet dissemination Topology map at each node Route computation using Dijkstra’s algorithm

OSPF advertisement carries one entry per neighbor router

Advertisements disseminated to entire AS (via flooding) Carried in OSPF messages directly over IP (rather than

TCP or UDP

Network Layer 4-36

OSPF “advanced” features (not in RIP)

Security: all OSPF messages authenticated (to prevent malicious intrusion)

Multiple same-cost paths allowed (only one path in RIP)

For each link, multiple cost metrics for different TOS (e.g., satellite link cost set “low” for best effort; high for real time)

Integrated uni- and multicast support: Multicast OSPF (MOSPF) uses same topology

data base as OSPF Hierarchical OSPF in large domains.

Network Layer 4-37

Hierarchical OSPF

Network Layer 4-38

Hierarchical OSPF

Two-level hierarchy: local area, backbone. Link-state advertisements only in area each nodes has detailed area topology; only know

direction (shortest path) to nets in other areas. Area border routers: “summarize” distances to

nets in own area, advertise to other Area Border routers.

Backbone routers: run OSPF routing limited to backbone.

Boundary routers: connect to other AS’s.

Network Layer 4-39

Internet inter-AS routing: BGP

BGP (Border Gateway Protocol): the de facto standard

BGP provides each AS a means to:1. Obtain subnet reachability information from

neighboring ASs.2. Propagate the reachability information to all

routers internal to the AS.3. Determine “good” routes to subnets based

on reachability information and policy. Allows a subnet to advertise its

existence to rest of the Internet: “I am here”

Network Layer 4-40

Anonymous System(AS)

Subnet 1.2

Subnet 1.4 Subnet

2.3

Subnet 2.2

Subnet 2.4

Subnet 2.1

Subnet 1.3

R1

R2R3

R4

R5

R6

R7

R8

Autonomous System 1 Autonomous System 2

LEGEND: Interior gateway protocol

Exterior gateway protocol

Subnet 1.1

Network Layer 4-41

Autonomous System (AS) AS

“ a set of routers under a single technical administration, using an Interior Gateway Protocol and common metrics to route packets within the AS and using an Exterior Gateway Protocol to route packets to other ASs” (RFC 1771)

Today AS may use more than one IGP, with potentially several sets of metrics.

The autonomous system designator is a 16-bit number, with a range of 1 to 65535. A range of AS numbers, 64512 through 65530, is reserved for private use, much like the private IP addresses.

Network Layer 4-42

Path attributes & BGP routes

When advertising a prefix, advert includes BGP attributes. prefix + attributes = “route”

Two important attributes: AS-PATH: contains the ASs through which the advert

for the prefix passed: AS 67 AS 17 NEXT-HOP: Indicates the specific internal-AS router

to next-hop AS. (There may be multiple links from current AS to next-hop-AS.)

When gateway router receives route advert, uses import policy to accept/decline.

Network Layer 4-43

BGP messages

BGP messages exchanged using TCP. BGP messages:

OPEN: opens TCP connection to peer and authenticates sender

UPDATE: advertises new path (or withdraws old)

KEEPALIVE keeps connection alive in absence of UPDATES; also ACKs OPEN request

NOTIFICATION: reports errors in previous msg; also used to close connection

Network Layer 4-44

AS 1

AS 5

AS 2

AS 3

1

1

BGP Example

R1

R2

R3R51

R52

AS 4R4

N3

N2

N1

Network Layer 4-45

BGP Example R1

Obtain the network information in AS1 by exchanging the IGP routing messages between routers in AS1.

R1 sends BGP UPDATE message to all BGP neighbors. The UPDATE message includes:

• AS_Path: {AS1}• Next_Hop: {R1’s IP address}• NLRI: {N1, N2, N3}

R2, R3 R2 and R3 that received the UPDATE message can

know that the networks in AS1 can be reached via R1.

Network Layer 4-46

AS 1

AS 5

AS 2

AS 3

2

2

BGP Example

R1

R2

R3R51

R52

AS 4R4

N3

N2

N1

Network Layer 4-47

BGP Example

R2 sends to R51 the UPDATE message that includes the following: AS_Path: {AS1, AS2} Next_Hop : {R2’s IP address} NLRI: {N1, N2, N3}

R3 sends to R4 the UPDATE message that includes the following: AS_Path: {AS1, AS3} Next_Hop : {R3’s IP address} NLRI: {N1, N2, N3}

Network Layer 4-48

AS 1

AS 5

AS 2

AS 3

3

BGP Example

R1

R2

R3R51

R52

AS 4R4

N3

N2

N1

Network Layer 4-49

BGP Example

R4 sends to R52 the UPDATE message that includes the following: AS_Path: {AS1, AS3, AS4} Next_Hop : {R3’s IP address} NLRI: {N1, N2, N3}

Network Layer 4-50

AS 1

AS 5

AS 2

AS 3

BGP Example

R1

R2

R3R51

R52

AS 4R4

N3

N2

N1

Network Layer 4-51

BGP Example R51 and R52

R51 that received the UPDATE message from R2 can know:• The networks in AS1 can be reached via R2 • The path is {AS2, AS1}.

R52 that received the UPDATE message from R4 can know:• The networks in AS1 can be reached via R4• The path is {AS4, AS3, AS1}.

R51 and R52 exchange BGP UPDATE messages. They will choose the better path.

Choosing the path, they refer to AS-path list in the UPDATE message to find out:

• Route loop(loop prevention)• Path selection based on the shortest path(shortest path vector)• Path selection based on AS policy(policy routing)

Network Layer 4-52

BGP route selection

Router may learn about more than 1 route to some prefix. Router must select route.

Elimination rules:1. Local preference value attribute: policy

decision2. Shortest AS-PATH 3. Closest NEXT-HOP router: hot potato

routing4. Additional criteria

Network Layer 4-53

BGP routing policy

Figure 4.5-BGPnew: a simple BGP scenario

A

B

C

W X

Y

legend:

customer network:

provider network

A,B,C are provider networks X,W,Y are customer (of provider networks) X is dual-homed: attached to two networks

X does not want to route from B via X to C .. so X will not advertise to B a route to C

Network Layer 4-54

BGP routing policy (2)

Figure 4.5-BGPnew: a simple BGP scenario

A

B

C

W X

Y

legend:

customer network:

provider network

A advertises to B the path AW B advertises to X the path BAW Should B advertise to C the path BAW?

No way! B gets no “revenue” for routing CBAW since neither W nor C are B’s customers

B wants to force C to route to w via A B wants to route only to/from its customers!

Network Layer 4-55

Why different Intra- and Inter-AS routing ?

Policy: Inter-AS: admin wants control over how its traffic

routed, who routes through its net. Intra-AS: single admin, so no policy decisions

needed

Scale: hierarchical routing saves table size, reduced

update trafficPerformance: Intra-AS: can focus on performance Inter-AS: policy may dominate over performance