ebook img

ATM switches PDF

335 Pages·1997·1.165 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview ATM switches

ATM Switches Edwin R. Coover Artech House Boston • London LibraryofCongressCataloging-in-PublicationData Coover,EdwinR.,1942– ATMswitches/EdwinR.Coover. p. cm. Includesbibliographicalreferencesandindex. ISBN0-89006-783-X(alk.paper) 1.Asynchronoustransfermode. I.Title. TK5105.35.C66 1997 004.6’6—dc21 97-12575 CIP BritishLibraryCataloguinginPublicationData Coover,EdwinR. ATMswitches 1. Asynchronoustransfermode I.Title. 621.3’8216 ISBN0-89006-783-X CoverdesignbyDarrellJudd © 1997ARTECHHOUSE,INC. 685CantonStreet Norwood,MA02062 All rights reserved. Printed and bound in the UnitedStates of America. Nopart of this book may be reproduced or utilized in any form or by any means, elec- tronic or mechanical, including photocopying, recording, or by any information storageandretrievalsystem,withoutpermissioninwritingfromthepublisher. All terms mentioned in this book that are known to be trademarks or service markshavebeenappropriatelycapitalized.ArtechHousecannotattesttotheac- curacy of this information. Use of a term in this book shouldnotberegardedas affectingthevalidityofanytrademarkorservicemark. InternationalStandardBookNumber:0-89006-783-X LibraryofCongressCatalogCardNumber:97-12575 10987654321 Contents Preface ix Acknowledgments xx Chapter1 Introduction 1 1.1 Why ATM? 1 1.2 TheState of theStandard 3 1.3 ATMConstituenciesToday 5 1.4 Sources,Sinks,andthe Future 8 Chapter2 TheATM Standard 11 2.1 Cells 12 2.2 A LayeredProtocol 12 2.3 ClassesofService 13 2.4 Signaling 15 2.5 CallsandConnections 15 2.6 Connection-oriented 16 2.7 Traffic Types 17 2.8 ATMAdaptationLayers (AAL) 17 2.8.1 AAL1 18 2.8.2 AAL2 19 2.8.3 AAL3/4 19 2.8.4 AAL5 19 2.9 Addressing 20 2.10 NetworkInterfaces 21 2.11 Summary 23 Chapter3 TheTechnicalChallengesof High-SpeedSwitchingI 25 3.1 InteroperabilityWithLegacySystems 25 3.2 Makinga Connection 28 v vi ATMSwitches 3.3 Multiplexing andDemultiplexing 31 3.4 Error Control 34 3.5 Latency 36 3.6 Data Packaging 36 3.7 Performance 37 3.8 Sequencing 38 3.9 Buffers andBuffering 39 3.10 Queues 40 3.11 CongestionControlandFlow Control 41 3.12 Summary 47 Chapter4 TheTechnicalChallengesof High-SpeedSwitchingII 49 4.1 Compression 49 4.2 Cost 50 4.3 Management 51 4.4 ProtocolsandProtocol Processing 54 4.5 Classof Service 56 4.6 Security 58 4.7 Fault Tolerance 60 4.8 Chargeback 62 4.9 SynchronizationandTiming 63 4.10 Summary 64 Chapter5 CommercialATMProducts 67 5.1 Prestandard 67 5.2 NetworkInterface Cards(NICs) 68 5.3 Plain Creative Marketing 69 5.4 TheManagement Shortfall 69 5.5 YourMileageWill Vary 76 5.6 TheLegacyoftheLegacy 77 5.7 Presentism 77 5.8 Today’sSwitchingVendors 78 5.9 ATMSwitch Overview 81 5.10 Summary 81 Chapter6 ATMLANSwitches 89 6.1 FourApproachesto Legacy LANs 92 6.2 ForeSystems 93 6.3 TheVirtualLANApproach 94 6.4 THeASX-200WG (For Workgroup) ATMSwitch 98 6.5 ForeThought 101 6.6 CiscoCorporation 107 6.7 Cisco’sLightstream1010WorkgroupSwitch 109 vii 6.8 CiscoHardware 109 6.9 CiscoSoftware 110 6.10 Summary 112 Chapter7 ATMAccessor“Edge”Switches 115 7.1 CascadeCommunicationsCorporation 116 7.2 Cascade’sB-STDX9000SwitchFamily 117 7.3 B-STDXATMMode 120 7.4 B-STDXFRMode 121 7.5 B-STDXSMDSMode 122 7.6 Newbridge Networks Corporation 123 7.7 Newbridge 36150AccessSwitch 125 7.8 Summary 129 Chapter8 ATMWANAccessProducts 131 8.1 ATMCSUs/DSUs 132 8.2 ATMInverse Multiplexers(AIMs) 132 8.3 ATMService Multiplexers 133 8.4 ADCKentrox 133 8.5 ADCKentroxDatasmartT3 ATMData Service UNIT(ADSU) 144 8.6 ADCKentroxT1/E1 ATMAccessConcentrator (AAC-1) 144 8.7 ADCKentroxMultismartSNMP Agency(MSA) 145 8.8 Digital Link Corporation 147 8.9 DL3200SMDS/ATMDigital Service Interface 148 8.10 W/ATMPremisewayATMService Access Multiplexer 149 8.11 Summary 152 Chapter9 Large ATMSwitches 155 9.1 NortelandtheMagellanFamily 156 9.2 Voice Networking 157 9.3 MagellanMultiplePriority System(MPS) 160 9.4 Nortel’sConcorde ATMSwitch 162 9.5 FujitsuCorporation 165 9.6 Fujitsu’sFETEX-150 ESP 166 9.7 DesignObjectives 166 9.8 TechnicalDescription 168 9.9 Marketing theFETEX-150 170 9.10 Video Server/the North Carolina Information Highway (NCIH) 170 9.11 Frame Relay/NYNEX 172 viii ATMSwitches 9.12 Cell Relay/BayNetworks 173 9.13 Summary 173 Chapter10 TheATMPrognosis 177 10.1 The“AnchorageAccord” 177 10.2 “Fast”Ethernet andthePromise of a Gigabit Ethernet 178 10.3 Standards,theForum, andthe Marketplace 179 10.4 TheLonger Term 180 Bibliography 183 Glossary 239 About theAuthor 325 Index 327 Preface Preface From its inception, this volume has followed a twisted road. Originally conceived as a book on one of several promising new broadband tech- nologies, the appearance of at least a dozen books on the emerging asyn- chronous transfer mode (ATM) standard clearly obviated that need. At the same time, the introduction of commercial ATM switches and com- ponents, and their peculiar interpretations and implementations of the standard, prompted a reappraisal. What is going on here? Has ATM, the much-touted “unification” technology, a single way of handling voice, video, and computer data across conventional local area network/metro- politan area networks/wide area network (LAN/MAN/WAN) boundaries, beenhighjacked? Itwouldseemso.First,anumberofsmallstartupcompaniessuchas FORE and Lightstream charged on the scene with “prestandard” LAN products. Their motto seems to be, “data can’t wait!”They, in turn,have been followed by seemingly every company that produces time division multiplexers (TDMs), routers, LAN hubs, packet orframe relay switches, or central offices switches. The credo is that ATM is “The Next Big Thing.” The level of sheer hype on ATM makes even Microsoft’s Windows 95launchlook modest.DoesATMpresent“IntegrationTechnology AND End-to-EndTechnology?”Isthehigh-speedATMnetworkthe“Backplane of the Metacomputer?” Will ATM result in the “Death of Distance?” Not very likely. Indeed, if anything, there is a bit of embarrassment on just where allthis speed andeleganttechnology willbeputtouse.In thelan- guageoftheindustry flak,thereis noacknowledged“killerapp”—a tech- nology-enabled breakthrough application. Instead, there is a host of contenders or perhaps pretenders: Lotus’ (now IBM) Notes, desktop videoconferencing, distance learning,and various applicationsinvolving compressedvideo. ix x ATMSwitches TheATMmarketisagoldrush,completewithChaplinesqeaspects, but mostly it is messy and wild. It is the most exciting technology since the Ethernet. Were it not for the Ethernet precedent, one would be temptedtoobservethattherearetoomanycompaniesforanyonetomake any money; indeed, it may be just the beginning. Nonetheless, like most goldrushes,manywillreturn(relatively)empty-handed.Lightstreamwas bought by router Goliath Cisco Systems, who subsequently gobbled up StrataCom.BBN’sATMswitchbecameAgileNetworks’ATMizer.Startup Centillion Networks was acquired by Bay Networks, itself a merger of WellfleetandSynOptics.Andjustbeforethisbookwenttopress,Cascade wasacquiredbyAscend, afast-risingvendorofremote-accessproducts. From the commercial product point of view—the view of this work—it is the exciting story of when a promising technology meets the magicofthemarketplace,andtheunexpectedhappens. Acknowledgments I have sometimes seen writing a book as a centripetal act in a centrifugal universe. Over the four years I have been researching and writing this one, the centripetal act has been much aided by the comments of col- leagues (or ex-colleagues) at the MITRE Corporation. Rick Tibbs, Rob Rice, Randy Mitchell, David Koester, Satish Chandra, Susmit Patel, Rich Verjinski, Doug Bamford, and Tom Ramsay have influenced me on a number of communications issues. Janet Park produced the figures and tables, and they reflect her considerable skill and care. The staff at the MITRElibrary,includingsomenowdeparted,providedtheresourcesand an environment that enabled sustained investigation. They included Sherri Lieberman, Vi Jefferson, Hugh Keleher, John Huddy, Win Kutz, SheliaChatterjee,JeanWatson,andBelindaParker. Mark Walsh,acquisi- tions editor at Artech House, has been both supportive and patient. Artech staff Lisa Tomasello, Pamela Ahl, Beverly Cutter, Matt Gambino, Kimberly Collignon, Julie Lovelock, Roberta de Avila, Alexia Rosoff, Darrell Judd,Kate Feininger,andJudiStoneprovidedtimely adviceona large span of issues and, in the process, averted some real howlers. Fi- nally, I owe a considerable debt to an unknown outside reader who not only caught a number of embarrassing errors, but whose witty marginal comments backed me off from a number of judgments that appeared, in retrospect,excessivelyexcathedra. xi Introduction 1 Introduction 1.1 WHY ATM? One can identify three developments of the 1980s that impelled the de- velopmentofhigh-speedswitching. The first was the continuing mismatch between computer channel speeds and the public transmission network. Twenty years ago it was “MIPS talking to Kbps;” now, it is close to “GIPS talking to Mbps.” Al- thoughthetransmissionnetworkscould,bymeansofseveraltiersofmul- tiplexingand demultiplexing,partiallyaccommodate thistorrentofzeros and ones, it was neither economical nor did it make sense intuitively to multiplex and demultiplex bit streams whose end use was a high-band- width device. To provide an example, a Cray supercomputer with a 100-Mbyte/s channel (800 Mbps) driving a graphics application would havehadtobemuxedintomultipleDS-3(45Mbps)streamsandreassem- bled. The costs of this sort of mismatch simply ensured that for such ap- plications, the cheapest routewas usually to move the applicationto the supercomputerorthesupercomputertotheapplication. Theseconddevelopmentwasthewidespreaduseofgraphicalrepre- sentation.Widelyusedas ameans ofvisualizationand/ordatareduction, the bit streams (even when compressed) again starkly outlined the mis- match between many graphical applications and an economic means to remotely distributethem. Forexample, a megapixel display(800×1200) using just 16-bit color graphics and employing only 15 updates per sec- ond generates a 230.4-Mbps bit stream. Although this 230.4-Mbps bit stream may (depending on the application’s requirements) be satisfacto- rily compressible ontoa DS-1(1.544 Mbps) circuit,a leased DS-1facility was often unjustifiable for a single high-bandwidth application; further, the obvious lower cost alternative, dial-up DS-1, is not ubiquitously available. 1

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.