Switched Local Area Network Design Switched LAN...

09_Progettazione_L2 - 1 Copyright 2002 - M. Baldi - P. Nicoletti: see page 2

Switched Local Area Network Design

-

Switched LAN

Switched Local Area Switched Local Area Network DesignNetwork Design

--

SwitchedSwitched LANLAN

Mario Baldi

Politecnico di Torinohttp://staff.polito.it/mario.baldi

Mario Baldi

Politecnico di Torinohttp://staff.polito.it/mario.baldi

Based on chapter 9 of: M. Baldi, P. Nicoletti, Switched LAN, McGraw-Hill, 2002, ISBN 88-386-3426-2


Copyright NoticeThis set of transparencies, hereinafter referred to as slides, is protected by copyright laws and provisions of International Treaties. The title and copyright regarding the slides (including, but not limitedto, each and every image, photography, animation, video, audio, music and text) are property of the authors specified on page 1.

The slides may be reproduced and used freely by research institutes, schools and Universities for non-profit, institutional purposes. In suchcases, no authorization is requested.

Any total or partial use or reproduction (including, but not limited to, reproduction on magnetic media, computer networks, and printedreproduction) is forbidden, unless explicitly authorized by the authorsby means of written license.

Information included in these slides is deemed as accurate at the date of publication. Such information is supplied for merely educational purposes and may not be used in designing systems, products, networks, etc. In any case, these slides are subject to changeswithout any previous notice. The authors do not assume anyresponsibility for the contents of these slides (including, but notlimited to, accuracy, completeness, enforceability, updated-ness of information hereinafter provided).

In any case, accordance with information hereinafter included mustnot be declared.

In any case, this copyright notice must never be removed and must bereported even in partial uses.


Topics

Design criteria

Design based on repeaters and switches

Fault tolerance

Design of fibre-basedbackbones


Design criteriaDesign criteria


Design criteriaDesign criteria

PerformancesPerformances

ReliabilityReliability

SecuritySecurity

Fault toleranceFault tolerance


Selection of network devices Selection of network devices

ReliabilityReliability

Observance of standard specifications

Observance of standard specifications

Cabling Cabling

Particular attention is needed forfibre-based backbones

Particular attention is needed forfibre-based backbones

AttenuationAttenuation


Transmission bandwidth Transmission bandwidth

Network resourcesNetwork resources

Packet processing capabilities of network devices

Packet processing capabilities of network devices


Depends onDepends on

Time Time

Different moments Different moments

Hours Hours

Resource employmentResource employment

Days of the week Days of the week

Special events Special events

Traffic typology Traffic typology


Client-server trafficClient-server traffic

Server with high-speed connections(for example 1 Gb/s)

Server with high-speed connections(for example 1 Gb/s)

Clients with lower speed connections

(for example100 Mb/s)

Clients with lower speed connections

(for example100 Mb/s)

Concentratedtraffic flowsConcentratedtraffic flows


SwitchSwitch

Crossed traffic flowsCrossed traffic flows

Peer-to-peer trafficPeer-to-peer traffic


Network traffic surveyNetwork traffic survey

Time Time

Resourcesemployment

Resourcesemployment

Functionality of network analysys monitors

Functionality of network analysys monitors

0 0

60%

50%

40%

30%

20%

10%

5%

0%

60%

50%

40%

30%

20%

10%

5%

0%


Simple to realize Simple to realize

Inexpensive Inexpensive

DimensioningDimensioning

The most common approach is to overdimension the network



No traffic engineering No traffic engineering

No resource reservation No resource reservation

Simple to manage Simple to manage

DimensioningDimensioning




Design guidelines

Device types and topologies

Design guidelines

Device types and topologies


BridgeBridge

Shared channelShared channel

Once it was also physical structureOnce it was also physical structure

Logical structureLogical structure


Based on point to point linksBased on point to point links

HubMultiportrepeater

HubMultiportrepeater

Physical structurePhysical structure


Star typologyStar typology

if necessary, it can be hierarchicalif necessary, it

can be hierarchical


Suitable linksSuitable links1 Gb/s1 Gb/s

100Mb/s

100Mb/s

10 Mb/s10 Mb/s


CoreBackbone

CoreBackbone

DistributionAggregationDistributionAggregation

AccessAccess

Normally on 3 levelsNormally on 3 levels


DatacenterBuilding cabinet

DatacenterBuilding cabinet

Horizontal wiringHorizontal wiring

VerticalwiringVerticalwiring

FloorcabinetFloor

cabinet

Normally on 3 levelsNormally on 3 levels


In a buildingIn a building

DatacenterDatacenter

Floor cabinetFloor cabinet

HorizontalwiringHorizontalwiring

Vertical wiringVertical wiring


Design for faulttoleranceDesign for faulttolerance


LinksLinks


The network can operate also when facing one or more failures

The network can operate also when facing one or more failures

DevicesDevices


Device partsDevice parts


InterfaceInterface

Power supplierPower supplier

Thanks toPietro Nicoletti

Thanks toPietro Nicoletti

The network can operate also whenfacing one or more failures

The network can operate also whenfacing one or more failures


How to realize itHow to realize it

Adding redundancy on critical elementsAdding redundancy on critical elements

InterfaceInterface

PortPort

ProcessorProcessor

Power supplierPower supplier

LinkLink

Device (for example a switch)Device (for example a switch)


Where to introduce redundancy?Where to introduce redundancy?

Processors, power suppliers, NICs Processors, power suppliers, NICs

At device levelAt device level

Parallel interfaces, redundant ports Parallel interfaces, redundant ports

At interface levelAt interface level

Alternative paths Alternative paths

At network levelAt network level Also combining all of these


How much redundancy?How much redundancy?

Each element has associateda fault probability and a costEach element has associateda fault probability and a cost

Rendundancy means to addnew elements

Rendundancy means to addnew elements

By increasingBy increasing

The probability of a faultThe probability of a fault

The cost of the networkThe cost of the network


Costs Costs

Real tolerance needs Real tolerance needs

The fault tolerant solution is a compromise among

The fault tolerant solution is a compromise among

EssentiallyEssentially

Quality (reliability) of elements Quality (reliability) of elements


A golden ruleA golden rule

The fault tolerant solution mustbe as simple as possible and use the lowest number of redundant elements required to guarantee a path that is alternative to

the faulty one

The fault tolerant solution mustbe as simple as possible and use the lowest number of redundant elements required to guarantee a path that is alternative to

the faulty one


Uninterruptible Power Supply systems for important devicesUninterruptible Power Supply systems for important devices

Good practicesGood practices

For example star center For example star center

Systems for the management and control of the network for revealing anomalies and faults

Systems for the management and control of the network for revealing anomalies and faults

Redundant devices must have anindependent power supply

Redundant devices must have anindependent power supply


Ring: good quality/price ratio

Ring: good quality/price ratio

Redundant linksRedundant links

Physical paths (channels) Physical paths (channels)

Avoid the sharing of critical elementsAvoid the sharing of critical elements

Optical fibre is preferableOptical fibre is preferable

BackboneBackbone

Link ends (panels) Link ends (panels)


Redundancy of the star centerRedundancy of the star center


Redundancy and ProtocolsRedundancy and Protocols

Alternative paths areDifficult for switches

Alternative paths areDifficult for switches

Spanning tree protocolIEEE 802.1d

Spanning tree protocolIEEE 802.1d


Spanning tree protocolSpanning tree protocol

Choice of the root bridgeChoice of the root bridge

Bridge Protocol Data Unit (BPDU)Bridge Protocol Data Unit (BPDU)

Selection of the root portsSelection of the root ports

Selection of the designated portsSelection of the designated ports


Spanning tree reconfigurationSpanning tree reconfiguration

Fault reactionFault reaction

Needed time: 50 secNeeded time: 50 sec


It can be managed by modifyingthe time-out

It can be managed by modifyingthe time-out

Maximum 7 cascading bridgesMaximum 7 cascading bridges

Underused resources Underused resources

Single routing treeSingle routing tree

Slow reconfigurationSlow reconfiguration

Drawbacks of the spanning tree protocolDrawbacks of the spanning tree protocol


Reconfiguration time: less than 1 s Reconfiguration time: less than 1 s

Rapid spanning tree protocol IEEE 802.1wRapid spanning tree protocol IEEE 802.1w

Managing the limits of the spanning tree protocolManaging the limits of the spanning tree protocol

Overcoming dimension limits Overcoming dimension limits

Multiple spanning tree protocol IEEE 802.1sMultiple spanning tree protocol IEEE 802.1s

Hierarchy of trees Hierarchy of trees


Blocked ports Blocked ports

Designing the details of the treeDesigning the details of the tree

Spanning tree protocol designSpanning tree protocol design

It replaces the root bridge facing a failure

It replaces the root bridge facing a failure

Chose the backup root bridgeChose the backup root bridge

Chose the root bridgeChose the root bridge


In the case: time-out tuningIn the case: time-out tuning

Study of the optimal tree in normal conditions

Study of the optimal tree in normal conditions

At leastAt least

Study of the tree during failures

Study of the tree during failures

Maximum diameter: 7 switches Maximum diameter: 7 switches

Spanning tree protocol designSpanning tree protocol design


Network and spanning treeNetwork and spanning treeRoot bridgeRoot bridge

Root portRoot port

Designated portDesignated port

Blocking portBlocking port

LEGENDALEGENDA

S1S1--PP

2457624576S1S1--SS

2867228672

S2S2

3276832768

S3S3

3276832768

S4S4

3276832768

S5S5

3276832768

S6S6

3276832768

S7S7

3276832768

BridgepriorityBridgepriority


Active topology without faultsActive topology without faults

S1S1--SS

2867228672

S2S2

3276832768

S3S3

3276832768

S4S4

3276832768

S5S5

3276832768

S6S6

3276832768

S1S1--PP

2457624576

S7S7

3276832768


Reaction to faultsReaction to faults

S1S1--PP

2457624576S1S1--SS

2867228672

S2S2

3276832768

S3S3

3276832768

S4S4

3276832768

S5S5

3276832768

S6S6

3276832768

S7S7

3276832768Root portRoot port



LEGENDALEGENDA

FaultFault



S1S1--PP

2457624576S1S1--SS

2867228672

S2S2

3276832768

S3S3

3276832768

S4S4

3276832768

S5S5

3276832768

S6S6

3276832768

Fault

Fault

S7S7

3276832768Root portRoot port



LEGENDALEGENDA



S1S1--PP

2457624576S1S1--SS

2867228672

S2S2

3276832768

S3S3

3276832768

S4S4

3276832768

S5S5

3276832768

S6S6

3276832768

S7S7

3276832768

Root bridgeRoot bridge

FaultFault

Root portRoot port



LEGENDALEGENDA


S1S1--PP

3276832768S1S1--SS

3276832768

S3S3

3276832768

S2S2

3276832768S5S5

3276832768S6S6

3276832768S4S4

3276832768S7S7

3276832768

Active topology with a bad configurationActive topology with a bad configuration

Root bridgeRoot bridge


SW-1 SW-2

1 2 3

4

3 2 1

4

A

B

C

D

E

F

B. Priority

24576B. Priority

286726 6

Multiple floor switches

Switched Local Area Network Design Switched LAN...

Documents

Transcript of Switched Local Area Network Design Switched LAN...