Internet inter-domain routing

LAN Design

1
Chapter 1
 LAN Design
Switched LAN Architecture

2
Chapter 1
 Switched LAN Architecture
• When building a LAN that satisfies the needs of a small or

medium-sized business, your plan is more likely to be

successful if a hierarchical design model is used.
 Divided into discrete layers.

 Each layer has a specific purpose.

 Becomes modular – maintenance, performance.

3
Chapter 1
 Switched LAN Architecture

4
Chapter 1
 Access Layer

• Interfaces with end devices.

• Routers, switches, bridges, wireless access points.
• Provides a means of connecting and controlling which devices are allowed to communicate
on the network.
5
Chapter 1
 Distribution Layer

• Aggregates (funnels) Access Layer traffic.

• Controls traffic flow with security or routing policies.
• Defines broadcast domains.
• Routing of VLANs (Virtual LANs).
6
Chapter 1
 Core Layer

• High speed backbone of the network.

• Must be highly available and redundant.
• Must be capable of quickly forwarding large amounts of data.
• Smaller networks – collapsed model (Core and Distribution).
7
Chapter 1
 Medium Sized Business

Logical
Layout

Physical
Layout

8
Chapter 1
 Benefits of a Hierarchical Network
• Benefits:
 Scalability
 Redundancy
 Performance
 Security
 Manageability
 Maintainability

9
Chapter 1
 Benefits of a Hierarchical Network
Scalability

Hierarchical Networks can be expanded easily.

10
Chapter 1
 Benefits of a Hierarchical Network
Redundancy

Redundancy at the core and distribution layers

11
ensure availability. Chapter 1
 Benefits of a Hierarchical Network
Performance

Link aggregation and high performance distribution and core

12
layer switches provide near-wire speed at all layers. Chapter 1
 Benefits of a Hierarchical Network
Security

Port security at the access layer and policies at the distribution

13
layer make the network more secure. Chapter 1
 Benefits of a Hierarchical Network
Manageability

Configurations

Additional Switch
Functionality
Rapid Recovery

Easier Troubleshooting

Consistency among switches at each layer makes

14
management more simple. Chapter 1
 Benefits of a Hierarchical Network
Maintainability

The modular design allows a network to scale easily without

15
becoming over-complicated or burdensome. Chapter 1
 Principles of Hierarchical Network
Design
• Just because a network is hierarchical, it doesn’t mean it’s
well designed.
 Network Diameter:
 The number of devices that a packet has to cross before it reaches its
destination.
 Bandwidth Aggregation:
 After the bandwidth requirements of the network are known, links
between specific switches can be aggregated or combined to provide
higher bandwidth.
 Redundancy:
 The practice of providing multiple paths to a destination or multiple
instances of a device.

16
Chapter 1
 Principles of Hierarchical Network Design

• Network Diameter:
 For PC1 to
communicate
with PC3, the
data must
traverse 6
intermediate
switches.
 In this case, the network diameter is 6.
 Each switch introduces some latency.
 In a hierarchical network, network diameter is always going to be a
predictable number of hops between the source and destination devices.

17
Chapter 1
 Principles of Hierarchical Network Design

• Bandwidth
Aggregation:
 Link aggregation
allows multiple
switch port
links to be
combined so
as to achieve higher throughput between switches.
 The determining factor is using link aggregation is the requirements
of the user applications.

18
Chapter 1
 Principles of Hierarchical Network Design

• Redundancy:
 Redundancy is
one part of
creating a
highly available
network.
 Multiple links
between
switches or multiple devices.
 It can get expensive and most likely will not be done on the
access layer because of the cost and variety of devices.
 It is feasible at the distribution and core layers.

19
Chapter 1
 What is a Converged Network?
• A Converged Network is one where voice and video
communications have been combined on a single data
network.
 Legacy Equipment:
 Until now, mainly feasible on large enterprise networks.

20
Chapter 1
 What is a Converged Network?
• Advanced Technology:
 More popular to medium and small sized businesses.
 Can be a difficult decision considering current investments in
technology.
 Benefit:
 Only one network to manage.

21
Chapter 1
 What is a Converged Network?
• New Options:
 You can now tie voice and video communications directly into
an employee's personal computer system.
 Software integrated on a PC eliminates an expensive handset.
 Add a webcam and video conference.

22
Chapter 1
 LAN Design
Matching Switches to
Specific LAN Functions

Traffic Flow Analysis

User Community Analysis

Data Stores and Data Servers Analysis

Topology Diagrams
Switch Features

23
Chapter 1
 Considerations for Network Switches
• Traffic Flow Analysis:

 The process of measuring the bandwidth usage on a network and

analyzing the data.

 Performance tuning.

 Capacity planning.

 Hardware improvement decisions.

24
Chapter 1
 Considerations for Network Switches
• User Community Analysis:
 The process of identifying various groupings of users and their
impact on network performance.

25
Chapter 1
 Considerations for Network Switches
• Data Stores and Data Servers Analysis:
 When analyzing traffic on a network, consider the location of
the data stores and data servers.
 Consider both client-server and server-server traffic.

26
Chapter 1
 Considerations for Network Switches
• Topology Diagram:
 A graphical representation of a network infrastructure.
 Switch connections with port numbers.
 Aggregated ports and redundant paths.
 Identify configuration by switch name.
 Could contain user information.

27
Chapter 1
 Switch Features
• Switch Form Factors:
 When selecting a switch, you need to decide between
 Fixed configuration or modular configuration.
 Stackable or non-stackable.
 The switch form factor (physical size) is important depending
upon where the switch will be installed.
 Wiring closet with limited space.
 Computer room with free standing racks.
 Shelf in a central area.

28
Chapter 1
 Switch Features
• Fixed Configuration Switches:
 Fixed in their configuration.
 You cannot add features or options to the switch beyond
those that originally came with the switch.

29
Chapter 1
 Switch Features
• Modular Switches:
 Offer more flexibility.
 Typically come with different sized chassis that allow for the
installation of different numbers of modular line cards.
 The line cards actually contain the ports.

30
Chapter 1
 Switch Features
• Stackable Switches:
 Interconnected using a special backplane cable that provides
high-bandwidth throughput between the switches (Cisco
StackWise).
 The stacked switches effectively operate as a single, larger
switch.
 Desirable when fault tolerance and bandwidth availability
are critical and a modular switch is too costly to implement.

31
Chapter 1
 Switch Performance
• When selecting a switch for the access, distribution, or core
layer, consider the ability of the switch to support:
 Port Density.
 Forwarding Rate.
 Bandwidth Aggregation Requirements.

32
Chapter 1
 Switch Performance
• Port Density:
 Port density is the number of ports available on a single switch.

24 Port

48 Port

Very high density.

33 Catalyst 6500 - 1,000 Ports
Chapter 1
 Switch Performance
• Forwarding Rate:
 Defines the processing capabilities of a switch by rating how
much data the switch can process per second.
 If the switch forwarding rate is too low, it cannot
accommodate full wire-speed communication across all of its
switch ports.
 A 48 port Gigabit switch is capable of switching 48 Gigabits of traffic.

34
Chapter 1
 Switch Performance
• Forwarding Rate:
 Access layer switches typically do not need to operate at full
wire speed because they are physically limited by their uplinks
to the distribution layer.
 Allows the use of:
 Less expensive, lower performing switches at the access layer.
 More expensive, higher performing switches at the distribution and
core layers, where the forwarding rate makes a bigger difference.

35
Chapter 1
 Switch Performance
• Link Aggregation:
 As part of bandwidth aggregation, you should determine if
there are enough ports on a switch to aggregate to support the
required bandwidth.
24 Port Gigabit
Up to 8 ports
bound together to
provide up to Single Port - 1 Gigabit
8 Gigabits. BOTTLE NECK

36
Chapter 1
 Switch Performance
• Power over Ethernet (PoE):
 Allows the switch to deliver power to a device over the existing
Ethernet cabling.

37 Adds considerable cost to the switch.

Chapter 1
 Switch Performance
• Layer 3 Functionality:
 Switches typically operate at Layer 2 of the OSI Model.

38
Chapter 1
 Switch Features – Hierarchical Network
• Access Layer Switch Features:

Port Security Link Aggregation

VLANs
FastEthernet/Gigabit
PoE

Quality of Service (QoS)

39
Chapter 1
 Switch Features – Hierarchical Network
• Distribution Layer Switch Features:

Layer 3 Support
Link Aggregation
High Forwarding Rate

Gigabit/10 Gigabit
Redundant Components

Quality of Service (QoS)

Security Policies
40
Chapter 1
 Switch Features – Hierarchical Network
• Core Layer Switch Features:

Layer 3 Support
Link Aggregation
Very High Forwarding Rate
Gigabit/10 Gigabit
Redundant Components

Quality of Service (QoS)

41
Chapter 1
 Switches – Small and Medium
Business (SMB)
• Cisco has seven switch product lines. Each product line offers
different characteristics and features, allowing you to find the
right switch to meet the functional requirements of your network.
Please
• The Cisco switch referlines
product to the
are: text or online
curriculum for details on each model.
 Catalyst Express 500
 Catalyst 2960
 Catalyst 3560
 Catalyst 3750
 Catalyst 4500
 Catalyst 4900
 Catalyst 6500

42
Chapter 1
 Switches – Small and Medium Business (SMB)
Summary Access Distribution Core
Bandwidth (Link) Aggregation u u u
FastEthernet/Gigabit Ethernet u
Gigabit Ethernet/10 Gigabit Ethernet u u
High Forwarding Rate u
Layer 3 Support u u
Port Security u
Power Over Ethernet (PoE) u
Quality of Service (QoS) u u u
Redundant Components u u
Security Policies/Access Control Lists u
Very High Forwarding Rate u
VLANs u
43
Chapter 1
 Routing Protocols

44
 Intra- and Inter-Domain Routing
 Autonomous System (AS): A group of networks under a
common administration and with common routing
policies. E.g. an ISP’s network
 Internet: A number of AS’s
 Intra-Domain Routing: Inside an AS
 IGP: Interior Gateway Protocol
 Inter-Domain Routing: Among AS’s
 EGP: Exterior Gateway Protocol

45
 Example

46
 Protocols to Discuss
 Routing Information Protocol (RIP)
 Open Shortest Path First (OSPF)
 Border Gateway Protocol (BGP)

47
 RIP
 Specified in RFC 2453
 Distance-vector routing
 Each router advertises its routing table to its neighbors
periodically
 Topology changes can be disseminated within a certain time
 Typical interval: 30s

48
 Scalability
 Max metric is 15, max network width is 15.
 Is it possible to increase the scalability by simply setting the
max metric to a larger number? E.g. 256.

49
 Scalability: Convergence
 Long convergence time may incur upon link failure
 Consider the failure of link B—D

50
 Scalability: Solutions
 Split Horizon: Do not advertise back to the source node. E.g.,
A  B, then B does not advertise back to A. But, it cannot
handle loops
 Triggered Updates: Instead of periodic advertisement,
immediate routing table dissemination is performed on
topology changes

51
 OSPF
 Intra Domain Routing Protocol
 RFC 2328
 Link-state routing
 The state of each link is propagated throughout the network
 Each router maintains a link state database, and has a picture of
the whole topology
 All routers have identical database, run identical algorithm and
have consistent routing tables

52
 Comparison with RIP
 Scalability: No limit on the network size
 Overhead: Update triggered only by link state change
 Convergence: Faster
 Multi-Path, Load-Balancing: a router may maintain multiple
paths to the same destination router

53
 OSPF: Link State Advertising
 Flooding
 Each LSA is flooded to a certain number of interfaces
 Each LSA is acknowledged
 Upon LSA transmission, a timers is started. Upon timeout
(without ACK), retransmission is performed

54
 Two-Layer Hierarchical Routing
 When network scales up, two-layer hierarchical routing
can be employed to reduce protocol traffic
 AS  multiple AREAs
 Internal router: inside an area
 Area border router: between areas, condenses the
routing information inside areas, forwards summaries
 Path cost: intra area + inter area + intra area

55
 Example
N1
Area 1 Area 0 3
R5 R10 AS2
N1
3
R1

R5 R10 AS2
4
1
N3 4
R9
R9
R6 R6
2
1
10
R8 N3 10
N2
R8
R2 Area 3
4

N6 5
6
Area 2 4
R7

N4 R4 3

1 N7
N7
1
R3
3

N4
N5 N6

N5

56
 BGP
 Inter-domain routing
 RFC 1171
 One domain does not have the internal information of another
domain
 Different domains may run different intra-domain routing
protocols
 The tasks of BGP
 Routing info exchange between domains
 Path cost definition
 Decision making of path selection

57
 Routing Information Exchange
 Specific TCP connections are established for BGP
 Four types of messages
 Open: confirm the session between two routers
 Update: advertise or withdraw routes
 Notification: error notification
 Keepalive: no routing information, keep connection

58
 Path Cost
 A number of path attributes, e.g.
 MED: multi-exit discriminator, small is good
 Local Preference: large is good
…

59
 Example

60
 Path Selection
 When a router is presented with multiple candidate paths,
path selection is performed based on the attributes
 The decision making is not defined in the protocol, network
administrators can make their own policies

61
 Path Selection: Cisco Example
 Prefer large weight
 Equal weight: large local preference
 Equal: the one from the local BGP speaker
 None: shortest AS length
 Equal: lowest origin type (IGP < EGP)
 Equal: lowest MED
 …

62
 Failure Recovery

63
 Failure Recovery
 Resume the interrupted services at the earliest possible time
 Approaches
 Restoration
 Protection
 Fast Reroute

64
 Restoration
 A failure is detected
 The failure is advertised throughout the entire network
 The topology is updated
 New paths are calculated
 Interrupted services are resumed

65
 Protection
 When the working paths (primary paths) are computed,
backup paths are computed at the same time
 Backup paths may not carry traffic during normal operation
 When a failure is detected, the affected traffic is switched to
the backup path(s) immediately

66
Reading Assignment

More explanations
Internet inter-domain routing
contents
 Border Gateway Protocol (BGP),
 BGP routing instability/slow convergence problems and
solutions
 BGP security problems and solutions
 New inter-domain routing protocol proposals.
INTRA- AND INTERDOMAIN ROUTING
 Routing inside an autonomous system is referred to

as intradomain routing.
 Routing between autonomous systems is referred

to as interdomain routing.
 Autonomous systems

70
 Popular routing protocols

71
 DISTANCE VECTOR ROUTING
 In distance vector routing, the least cost route
between any two nodes is the route with minimum
distance.
 In this protocol each node maintains a vector (table)
of minimum distances to every node
 The topics discussed in this section include:
 Initialization
 Sharing
 Updating
 When to Share
 Two-Node Loop Instability
 Three-Node Instability
 Distance vector routing tables

73
 Initialization of tables in distance vector routing

74 TCP/IP Protocol Suite

 Note:

In distance vector routing, each node

shares its routing table with its
immediate neighbors periodically and
when there is a change.

75
 Updating in distance vector routing

76
 Two-node instability

77
 Three-node instability

78
 RIP

 The Routing Information Protocol (RIP) is an

intradomain routing protocol used inside an autonomous
system.
 It is a very simple protocol based on distance vector
routing
 The topics discussed in this section include:
 RIP Message Format
 Requests and Responses
 Timers in RIP
 RIP Version 2
 Encapsulation
 Example of a domain using RIP

80
 RIP message format

81 TCP/IP Protocol Suite

 Request messages

82
 RIP version 2 format

83 TCP/IP Protocol Suite

 Authentication

84 TCP/IP Protocol Suite

 Note:

RIP uses the services of UDP on

well-known port 520.

85 TCP/IP Protocol Suite

LINK STATE ROUTING
 In link state routing, if each node in the domain
has the entire topology of the domain, the node can
use Dijkstra’s algorithm to build a routing table.
 The topics discussed in this section include
 Building Routing Tables
 Concept of link state routing

87
 Link state knowledge

88
 Dijkstra algorithm

89
 Example of formation of shortest path tree

90
 Table.1 Routing table for node A

91
 OSPF
 The Open Shortest Path First (OSPF) protocol is an
intradomain routing protocol based on link state
routing. Its domain is also an autonomous system.
 The topics discussed in this section include:
 Areas
 Metric
 Types of Links
 Graphical Representation
 OSPF Packets
 Link State Update Packet
 Other Packets
 Encapsulation
 Areas in an autonomous system

93
 Types of links

94
 Point-to-point link

95 TCP/IP Protocol Suite

 Transient link

96 TCP/IP Protocol Suite

 Stub link

97
 Example of an AS and its graphical representation
in OSPF

98
 Types of OSPF packets

99
 OSPF common header

100
 Link state update packet

101
 LSA general header

102
 Router link

103
 Router link LSA

104
 Table 2 Link types, link identification, and link data

105
 Note:

OSPF packets are encapsulated in IP

datagrams.

106 TCP/IP Protocol Suite

 BGP
 Border Gateway Protocol (BGP) is an interdomain routing
protocol using path vector routing.
 It first appeared in 1989 and has gone through four
versions.
 The topics discussed in this section include:
 Types of Autonomous Systems
 Path Attributes
 BGP Sessions
 External and Internal BGP
 Types of Packets
 Packet Format
 Encapsulation
 Internal and external BGP sessions

108 TCP/IP Protocol Suite

 Types of BGP messages

109 TCP/IP Protocol Suite

 BGP packet header

110 TCP/IP Protocol Suite

 Open message

111 TCP/IP Protocol Suite

 Update message

112 TCP/IP Protocol Suite

 Keepalive message

113 TCP/IP Protocol Suite

 Notification message

114 TCP/IP Protocol Suite

 Table 3 Error codes

115 TCP/IP Protocol Suite

 Note:

BGP uses the services of TCP

on port 179.

116 TCP/IP Protocol Suite