4-7 Network Layer.pptx

8/13/2019 4-7 Network Layer.pptx

1/109

Network Layer (Part 7)

Computer Networks

Tutun JuhanaTelecommunication EngineeringSchool of Electrical Engineering & Informatics

Institut Teknologi Bandung

4


2/109

UNICAST ROUTING PROTOCOLS(RIP, OSPF, AND BGP)


3/109

INTRODUCTION

3


4/109

Cost or Metric

A router is usually attached to several networkswhen it receives a packet, to which network

should it pass the packet?

The decision is based on optimization: Which ofthe available pathways is the optimum pathway?

What is the definition of the term optimum?

One approach is to assign a cost for passing

through a networkWe call this cost a metric High cost can be thought of as something bad;

low cost can be thought of something good

4


5/109

Static versus Dynamic Routing Tables

A static table is one with manual entries

A dynamic table, on the other hand, is one

that is updated automatically when there is

a change somewhere in the internet

5


6/109

Routing Protocol

A routing protocol is a combination of rules and

procedures that lets routers in the internet inform

each other of changes

It allows routers to share whatever they know

about the internet or their neighborhood The routing protocols also include procedures

for combining information received from other

routers

Routing protocols can be either an interior

protocol or an exterior protocol

An interior protocol handles intradomain routing

an exterior protocol handles interdomain routing 6


7/109

INTRA- AND INTER-DOMAINROUTING

7


8/109

8

a group of networks and routers under the authority of a single administration

intra-domain routing

inter-domain routing

AS Numbers is

assigned for each AS

Ex: ITBs ASN is

4796


9/109

9


10/109

DISTANCE VECTOR ROUTING

10


11/109

This method sees an AS, with all routersand networks, as a graph, a set of nodes

and lines (edges) connecting the nodes

A router normally be represented by a nodeA network be represented by a link

connecting two nodes

The graph theory used Bellman-Ford (alsocalled Ford-Fulkerson) algorithm to find

the shortest path between nodes in a

graph given the distance between nodes11


12/109

Bellman-Ford Algorithm

It looks circular

To solve the problem, we use iteration to create a shortest distance table(vector) for each node using the following steps:

1. The shortest distance and the cost between a node and itself is initialized to 0.

2. The shortest distance between a node and any other node is set to infinity. The cost between a

node and any other node should be given (can be infinity if the nodes are not connected)

3. The algorithm repeat as shown in Figure 11.4 until there is no more change in the shortest

distance vector.

12


13/109

13


14/109

Distance Vector Routing Algorithm

14


15/109


16/109

16


17/109

17


18/109

18


19/109

19


20/109

20


21/109

Count to Infinity

21


22/109

In distance vector routing, any decrease in cost(good news) propagates quickly, but any

increase in cost (bad news) propagates slowly

For a routing protocol to work properly, if a link is

broken (cost becomes infinity), every other

router should be aware of it immediately, but in

distance vector routing, this takes some time.

The problem is referred to as count to infinity takes several updates before the cost for a

broken link is recorded as infinity by all routers.

22


23/109

Example of count to infinity

Two-Node Loop

23


24/109

Another example24

A converged network

Source

http://technet.microsoft.com/en-s/library/cc940478.aspx


25/109

The solutions

25


26/109

Defining Infinity

The first obvioussolution to count to

infinity is to redefine

infinity to a smaller

number

26

Most implementations of the Distance VectorProtocol define 16as infinity

However, this means that distance vector cannot be

used in large systems The size of the network, in

each direction, can not exceed 15 hops


27/109

Split Horizon

27

Split horizon helps reduce convergence time by not

allowing routers to advertise networks in the

direction from which those networks were learned


28/109

28

Source

http://technet.microsoft.com/en-s/library/cc940478.aspx


29/109

Split Horizon and Poison Reverse

29

Iit announces all networks. However, those networks learned in a

given direction are announced with a hop count of 16, indicating

that the network is unreachable It avoids the Distance Vector

Protocol deleting the route because it has no news about it during a

certain time duration (timer)


30/109

30


31/109

Three-Node Instability

31


32/109

RIPRouting Information Protocol

32


33/109

RIP implements distance vector routing directly with someconsiderations

33


34/109

34


35/109

RIP Message Format

35


36/109

Requests and Responses

36


37/109

Request

A request message is sent by a router thathas just come up or by a router that has

some time-out entries.

37


38/109

Response

A response can be : Solicited: sent only in answer to a request.

It contains information about the destination

specified in the corresponding request Unsolicited: is sent periodically

every 30 seconds or

when there is a change in the routing table

The response is sometimes called an

update packet

38


39/109

39


40/109

Timers in RIP

40

controls the

advertising of

regular update

messages

Governs the validity of a

route. If the timer is

reached, route is

declared unreachable,

but does not immediatelypurge, instead, it

continues to advertise

the route with a metric

value of 16.

At the same time when a

route declared unreachable,

the garbage collection timer

is set to 120 s for that route.

When the count reacheszero, the route is purged

from the table.


41/109

41


42/109

RIP Version 2

It supports

Classless Addressing

Authentication

Multicastinguses the all-router multicast

address to send the RIP messages only to

RIP routers in the network.42


43/109

Encapsulation

43


44/109

LINK STATE ROUTING

44


45/109

45

Each node in the domain

has the entire topology

of the domainuse

Dijkstra algo r i thmto

build a routing table

The routing table for

each node is unique

because the calculations

are based on different

interpretations of the

topology


46/109

46

The whole topology can be compiled from the partial knowledge of each

node (it knows the state (type, condition, and cost) of its links)


47/109

BUILDING ROUTING TABLES

47


48/109

48


49/109

Creation of Link State Packet (LSP)

A link state packet(LSP) carries (among othershuge information)

1. The node identity

2. The list of links

3. A sequence number4. Age

The first two are needed to make the topology

The third facilitates flooding and distinguishesnew LSPs from old ones

The fourth prevents old LSPs from remaining in

the domain for a long time49


50/109


51/109

Flooding of LSPs

51

Formation of Shortest Path Tree:


52/109

Formation of Shortest Path Tree:Dijkstra Algorithm

After receiving all LSPs, each node willhave a copy of the whole topologynot

sufficient to find the shortest pathto every

other nodea shortest path treeisneeded

A shortest path treeis a tree in which the

path between the root and every othernode is the shortest

What we need for each node is a shortest

path tree with that node as the root52

Dijk l i h


53/109

Dijkstra algorithm

53


54/109

54


55/109

55


56/109

56

Calculation of Routing Table from


57/109

g

Shortest Path Tree

57


58/109

OSPFOpen Shortest Path First

58

A


59/109

Area

OSPF divides an autonomous system into

areas

An area is a collection of networks, hosts,

and routers all contained within anautonomous system

All networks inside an area must be

connected

59


60/109

Routers inside an area flood the area with

routing information

At the border of an area, special routers

called area border routerssummarize theinformation about the area and send it to

other areas

60


61/109

All of the areas inside an autonomoussystem must be connected to a special

area called the backbone area

The backboneserves as aprimary areaand

the other areas as secondary areas

This does not mean that the routers

within areas cannot be connected to

each other

The routers inside the backbone are

called the backbone routers

backbone router can also be an

area border router

Each area has an area identification

The area identification of the

backboneis zero

61


62/109

62


63/109

If, because of some problem, the

connectivity between a backbone and an

area is broken, a virtual linkbetween routers

must be created by the administration toallow continuity of the functions of the

backbone as the primary area

63

M t i


64/109

Metric

The OSPF protocol allows the

administrator to assign a cost, called the

metric, to each route

The metric can be based on a type ofservice (minimum delay, maximum

throughput, and so on)

As a matter of fact, a router can havemultiple routing tables, each based on a

different type of service64

T pes of Links


65/109

Types of Links

In OSPF terminology, a connection is called a link

65

Point to Point Link


66/109

Point-to-Point Link

There is no need to assign a network address to

this type of link

66

Transient Link


67/109

Transient Link A transient link is a network with several routers

attached to it

67

It is not efficienteach router needs

to advertise the neighborhood to four

other routers

It is not realisticthere is no single

network (link) between each pair of

routers (there is only one network

(not router)that serves as a crossroad

between all five routers)

To show that each router is connected to

every other router through one single

networkthe network itself is representedby a node

A network is not a machineit cannot

function as a router

So, one of the routers in the network takes

this responsibilityIt is assigned a dual

purposeit is a true routerand a

designated router


68/109

While there is a metric from each node to the designated

router, there is no metric from the designated router to

any other nodeWe can only assign a cost to a packet

that is passing through the network (We cannot charge

for this twice)

When a packet enters a network, we assign a cost; when a

packet leaves the network to go to the router, there is no charge

68

Stub Link


69/109

Stub Link

A special case of the transient network

The link is only onedirectional, from therouter to the network

69

Virtual Link


70/109

Virtual Link

When the link between two routers is

broken, the administration may create a

virtual link between them using a longer path

that probably goes through several routers

70

Graphical Representation


71/109

Graphical Representation

71

OSPF Packets


72/109

OSPF Packets

72

Common Header


73/109

Common Header

73


74/109


75/109


76/109

76


77/109

77


78/109

78

Network Link LSA


79/109

Network Link LSA

79


80/109

80


81/109

81


82/109

Summary Link to Network LSA


83/109

Summary Link to Network LSA

83

The summary link to network LSA is used by the area border router to

announce the existence of other networks outside the area


84/109

84

Summary Link to AS Boundary Router LSA


85/109

Summary Link to AS Boundary Router LSA

85


86/109

86

External Link LSA


87/109

External Link LSA

87

Other Packets


88/109

Other Packets

They are not used as LSAs, but areessential to the operation of OSPF

88

Hello Message


89/109

Hello Message

OSPF uses the hello message to create

neighborhood relationships and to test the

reachability of neighbors

89

Database Description Message


90/109

Database Description Message

90

Link State Request Packet


91/109

Link State Request Packet

91


92/109

Encapsulation


93/109

Encapsulation

93


94/109

PATH VECTOR ROUTING

94


95/109

Path vector routing is exterior routingprotocol proved to be useful for interdomain

or inter-AS routing

95

Reachability


96/109

Reachability

96

Routing Tables


97/109

Routing Tables

A path vector routing table for each router can

be created if ASs share their reachability list with

each other97

Loop Prevention


98/109

Loop Prevention

The instability of distance vector routing andthe creation of loops can be avoided in path

vector routing. When a router receives a

reachability information, it checks to see if itsautonomous system is in the path list to any

destination. If it is, looping is involved and

that network-path pair is discarded.

98

Aggregation


99/109

Aggregation

99

Policy Routing


100/109

y g

When a router receives a message, it cancheck the path. If one of the autonomous

systems listed in the path is against its

policy, it can ignore that path and thatdestination. It does not update its routing

table with this path, and it does not send this

message to its neighbors

100


101/109

BGPBorder Gateway ProtocolAn interdomain routing protocol

101

Types of Autonomous Systems


102/109

yp y

Stub AS

Multihomed AS

Transit AS

102

Stub AS


103/109

A stub AS has only one connection toanother AS

The hosts in the AS can send data traffic

to other Ass The hosts in the AS can receive data

coming from hosts in other Ass

Data traffic cannot pass through a stub AS A stub AS is either a source or a sink

103

Multihomed AS


104/109

A multihomed AS has more than oneconnection to other ASs, but it is still only

a source or sink for data traffic

It can receive data traffic from more thanone AS

It can send data traffic to more than one

AS, but there is no transient traffic

It does not allow data coming from one AS

and going to another AS to pass through

104

Transit AS


105/109

A transit AS is a multihomed AS that alsoallows transient traffic. Good examples of

transit ASs are national and international

ISPs (Internet backbones)

105


106/109

Path Attributes


107/109

Well-known attribute

One that every BGP router must recognize1. Well-known mandatory attribute

one that must appear in the description of a route

2. well-known discretionary attribute

one that must be recognized by each router, but is not required

to be included in every update message

Optional attribute

one that needs not be recognized by every router

1. optional transitive attribute

one that must be passed to the next router by the router that

has not implemented this attribute

2. optional nontransitive attribute

one that must be discarded if the receiving router has not implemented

it. 107


108/109

108

A session is a connection that is established between two BGP

routers only for the sake of exchanging routing information.


109/109

4-7 Network Layer.pptx

Documents

Transcript of 4-7 Network Layer.pptx