d7024e: distributed & mobile computing systems

D7024E: DISTRIBUTED & MOBILE COMPUTING SYSTEMS

Lecture 01: Introduction to distributed and mobile computing. Architectural models for mobile & distributed computing systems.

Coulouris Dollimore and Kindberg

Characterization of Distributed Systems

Coulouris, Dollimore and KindbergDistributed Systems: Concepts and Design

Edition 4, © Pearson Education 2005

MDCS: What we will cover ?MDCS: What we will cover ?

Introduction to Distributed SystemsBird’s eye view on mobile/pervasive/wireless y /p /computing

Why mobile revolution is inevitableWhy mobile revolution is inevitableAdvanced applications for mobile computing

A hit t d t f di t ib t d tArchitectures and components of distributed systemsConcepts and terminology

D7024E MDCS: Introduction

Recommended ReferencesRecommended References

Coulouris, Dollimore, Kindberg, “Distributed Systems”, 4th Ed, Addison-Wesley, 2005

A. Tanenbaum, Van Steen, “Distributed Systems”, Pearson, 2007A. Tanenbaum, Van Steen, Distributed Systems , Pearson, 2007

T.Imielinski, H. Korth “Mobile Computing”, Kluwer Academic publishers, 1996

G. Forman & J. Zahorjan “The challenges of mobile computing”, IEEE Computer, April 1994, pp.38-47

Deitel H Deitel P Nieto T Steinbuhler K "Wireless Internet & Mobile Deitel, H, Deitel, P., Nieto, T., Steinbuhler, K. Wireless Internet & Mobile business", Prentice Hall, 2001, ISBN 0-13-062226-5


What is a True Distributed SystemWhat is a True Distributed System

“A system that runs on a collection of machines that do not have shared memory, yet looks to its y, yusers like a single computer”

Eg Amoeba, Sprite, Chorus, CloudsSingle set of System calls on each nodeSingle set of System calls on each nodeAll machines run the same kernelMachines communicateEach kernel controls its own resources (?)

Definition of a Distributed System (2)Definition of a Distributed System (2)

Figure 1-1. A distributed system organized as middleware. The middleware layer extends over multiple machines, and offers each application the same interface

Tanenbaum & Van Steen, Distributed Systems: Principles and Paradigms, 2e, (c) 2007 Prentice-Hall, Inc. All rights reserved. 0-13-239227-5

same interface.

Figure 1.1A typical portion of the Internet

intranet

ISP

backbone

desktop computer:

satellite link

server:network link:

Instructor’s Guide for Coulouris, Dollimore and Kindberg Distributed Systems: Concepts and Design Edn. 4, © Pearson Education 2005

Figure 1.2A typical intranet

Desktopemail server pcomputers

print and other servers

email server

Web serverLocal areanetwork

email serverprint

th t f

File server

other servers

print

the rest of

router/firewall

the Internet


What is a Distributed Mobile Computing System

M bil i i i d M bil i i i d Mobile computing is associated Mobile computing is associated with mobility of with mobility of hardware, users, hardware, users, data and softwaredata and software in computer in computer ppapplications. applications. Specialised class of distributed Specialised class of distributed computing systems where some computing systems where some nodes can move in physical space, nodes can move in physical space, ad hocad hoc connecting/ disconnecting connecting/ disconnecting ad hocad hoc connecting/ disconnecting connecting/ disconnecting while remaining part of a while remaining part of a distributed system and perhaps distributed system and perhaps

ti i ti i l b l ti i ti i l b l participating in global participating in global computational activitiescomputational activities

Figure 1.3P bl d h dh ld d i i di ib d Portable and handheld devices in a distributed system

Internet

Host intranet Home intranetWAP Wireless LAN gateway

Mobilephone

LaptopPrinterCamera Host site


Figure 1.4Web servers and web browsers

BrowsersW b

www.google.comhttp://www.google.comlsearch?q=kindberg

Internet

BrowsersWeb servers

www.cdk3.nethttp://www.cdk3.net/

www.w3c.org

http://www.cdk3.net/

Protocols

A ti it ht l

http://www.w3c.org/Protocols/Activity.htmlFile system ofwww.w3c.org

Activity.html


Cluster Computing SystemsCluster Computing Systems

Figure 1-6. An example of a cluster computing system.


Grid Computing SystemsGrid Computing Systems


Figure 1-7. A layered architecture for grid computing systems.

Figure 1.5Computers in the Internet

Date Computers Web servers

1979, Dec. 188 0

1989, July 130,000 01999, July 56,218,000 5,560,8662003, Jan. 171,638,297 35,424,956


Figure 1.6Computers vs. Web servers in the Internet

Date Computers Web servers Percentagep g

1993, July 1,776,000 130 0.008

1995 July 6 642 000 23 500 0 41995, July 6,642,000 23,500 0.41997, July 19,540,000 1,203,096 61999 July 56 218 000 6 598 697 121999, July 56,218,000 6,598,697 12

2001, July 125,888,197 31,299,592 2542,298,37142,298,371


Th i l d t h llTh i l d t h llThe wireless data challengeThe wireless data challenge

To enable people to communicate/compute effortlessly, where they want, when they

want, without “Wires”

Mobile Computing: Problems and Assumptions

Location is dynamically changing informationCost of communication is asymmetricySevere power restrictions on mobile hostsLi it d t il bl bil h tLimited storage available on mobile hostsVolatility of information increases with the change of locationFrequent and foreseeable disconnectionsqSecurity issues due to mobility of hosts

Distributed Pervasive SystemsDistributed Pervasive Systems

Requirements for pervasive systemsRequirements for pervasive systems

• Embrace contextual changes.• Encourage ad hoc composition.Encourage ad hoc composition.• Recognize sharing as the default.


Pervasive computingPervasive computing

Th id th t t h l i i b d th l t tThe idea that technology is moving beyond the personal computer to everyday devices with embedded technology and connectivity as computing devices become progressively smaller and more powerful. Also called ubiquitous computing pervasive computing is the result ofcalled ubiquitous computing, pervasive computing is the result of computer technology advancing at exponential speeds ‐‐ a trend toward all man‐made and some natural products having hardware and software. Pervasive computing goes beyond the realm of personal computers: it isPervasive computing goes beyond the realm of personal computers: it is the idea that almost any device, from clothing to tools to appliances to cars to homes to the human body to your coffee mug, can be imbedded with chips to connect the device to an infinite network of other deviceswith chips to connect the device to an infinite network of other devices. The goal of pervasive computing, which combines current network technologies with wireless computing, voice recognition, Internet capability and artificial intelligence, is to create an environment wherecapability and artificial intelligence, is to create an environment where the connectivity of devices is embedded in such a way that the connectivity is unobtrusive and always available

http://www.webopedia.com/TERM/P/pervasive_computing.html

User ExpectationsUser Expectations

The user does not wish to beThe user does not wish to beconsidered exclusively as considered exclusively as Home Office

Ph

All-in-one device: convergence or divergence of applications?

considered exclusively as considered exclusively as a consumer.a consumer.

OfficeComputer

Phone

g g ppQuality of Experience: devices with intuitive, flexible, interactive and enabling interfaces, active guidance, personalised services, context awareness, price not an issue if the service proposition is compelling. Life assistant services: citizens wish to remain independent and mobile even in advanced Life assistant services: citizens wish to remain independent and mobile even in advanced years.Privacy safeguarding: requirements will increase in reaction to the growing possibilities of information interception and user profiling (location awareness and identity).Security technology: is demanded by users to cope with tracking services and Security technology: is demanded by users to cope with tracking services and surveillance systems

Challenges:Challenges:Bridging the gap from technological availability to actual market takeBridging the gap from technological availability to actual market take--upupBridging the gap from technological availability to actual market takeBridging the gap from technological availability to actual market take--upupCreating opportunities for userCreating opportunities for user--based innovationbased innovation

The following few slides are borrowed from J.DaSilva’s presentation at MDM’05, Cyprus, May, 2005

The world of Convergence

CECEMOBILEMOBILE BROADCASTBROADCASTBROADBANDBROADBAND

The world of Convergence

CECEPre-Recorded

ContentPersonal Media

MOBILEMOBILECommunication,Entertainment,

Pictures and Video

BROADCASTBROADCASTServices,

Entertainment

BROADBANDBROADBANDCommunication,Entertainment,

E-Business

Simple Networks Service Enabled Networks

Each island offers numerous Each island offers numerous services on distinct devicesservices on distinct devices

A converged archipelago offers A converged archipelago offers all services on all devicesall services on all devices

Challenges:Challenges:Providing endProviding end--toto--end service at minimal OPEX and CAPEXend service at minimal OPEX and CAPEXEnsuring the creation and management of convergenceEnsuring the creation and management of convergenceEnsuring the creation and management of convergenceEnsuring the creation and management of convergenceMoving from product supply to solution provisionMoving from product supply to solution provision

Services convergenceServices convergenceServices convergenceServices convergence

Anywhere

gamesgames

HDTVHDTVHome

gateway Any device

3

gamesgames

M bil

SatelliteSatellite

I i iMusic

AVprogram

ContentContent

3D TVComputerComputer Anytime

Mobileterminal

CableCableTerrestrialTerrestrial

Interactive serviceMultimedia

program

HDTVPC

Wired/wirelessWired/wireless3D A/V

DVB-H terminal

Networking convergenceNetworking convergence

DMBDVB-H (IP based core network) New radioNew radioNew radioNew radio

DTVDTV

Wire -linexDSL

Wire -linexDSL

CellularGSM-3G 4G WLAN

interfaceinterfaceinterfaceinterface

Satellite

T e le v is io n

B rid g eG a te w a y

WPAN

P D A Health sensor

C o m m . T o w e r

Home sensorL a p to p

P rin te r

V id e o

P D A

P D A

Cargo

Highway Nuclear

C o m m . T o w e r

C o m m . T o w e rP B X

C o m m . T o w e r

C o m m . T o w e r

Cargo sensor

River sensor

Bridge sensor

sensorOcean sensor

sensor

InteroperabilityInteroperability

At the Network/Device Level At the Network/Device Level Wireless/Mobile/Fixed/Cable/ISP/Broadcasting networks need to interoperate

At the Service/application LevelAt the Service/application LevelServices need to run across homogeneous or Services need to run across homogeneous or heterogeneous networks

At the Media/Content LevelAt the Media/Content LevelDiff t di f t t i tDifferent media formats must coexist

Challenges:Challenges:ggEnsuring smooth technological transitionsEnsuring smooth technological transitionsCreating opportunities for disruption and innovationCreating opportunities for disruption and innovationContributing to setting the right collaborative standardsContributing to setting the right collaborative standardsOptimising for innovation through a more agile IP regimeOptimising for innovation through a more agile IP regimeCoping with the trend towards the “patenting of software”Coping with the trend towards the “patenting of software”

Interoperability at Home

Internet, Music andInternet, Music and VoDVoD, Video, Video

PDAVideo walkman

Web TabletLaptop

Internet, Music and Internet, Music and TV/video everywhereTV/video everywhere

VoDVoD, Video , Video streaming, music streaming, music download, storagedownload, storage

Bluetooth andWi-Fi Phones

Users want their devices to work togetherand share contentHome Home

Media CentrePrinter

automation and automation and ControlControl

Media Centre

Game Console IP STB and PVRTeleworkTelework

Voice and video Voice and video conferencingconferencingTelemedicineTelemedicine

Challenges:Challenges:Ensuring content management interoperabilityEnsuring content management interoperabilitySt iki th i t b l b t i ht h ld d i htSt iki th i t b l b t i ht h ld d i htStriking the appropriate balance between right holders and consumers rightsStriking the appropriate balance between right holders and consumers rights

Explosion of Devices and DataExplosion of Devices and Data

Amount of data received or transmitted (in Petabytes/Day)

Information explosion and overload Information explosion and overload

Petabytes/Day)

1,000,000

1,200,000

IndustrialAutomobile

Number of communicating data Number of communicating data devices growing from 2.4 billion to 23 devices growing from 2.4 billion to 23 billion in 2008 and one trillion by 2012billion in 2008 and one trillion by 2012600,000

800,000

Automobile

Entertainment

billion in 2008 and one trillion by 2012billion in 2008 and one trillion by 2012

400,000

,Mobile

0

200,000

2003 2004 2005 2006 2007 2008

Computers

Challenges:Challenges:Designing and managing an information infrastructure where all devices communicate Designing and managing an information infrastructure where all devices communicate

ith d d t d thith d d t d th

2003 2004 2005 2006 2007 2008

with and understand one anotherwith and understand one anotherCreating an advanced digital ecoCreating an advanced digital eco--system for the agile enterprisesystem for the agile enterprise

The scale of networkingg

The wide area networks of The wide area networks of yesterday (eg: GSM)

> A Million nodes @ $50k

The Nomadic local area networks f d ( WiFi)of today (eg: WiFi)

> Millions of Nodes @ $100

The Sensor and Personal area network of tomorrow

@ $> Billions of Nodes @ $1Challenges:Challenges:

Removing social, geographical, economic and capacity Removing social, geographical, economic and capacity impediments impediments through the provision of through the provision of cost effective cost effective infrastructuresinfrastructures, allowing an “Always on” network existence., allowing an “Always on” network existence.Contributing to accrued facilities based competition.Contributing to accrued facilities based competition.

Ubiquitous computingUbiquitous computing

Suppose that you could get a 0 1$ “computer” Suppose that you could get a 0.1$ computer . What would you do with it?

Play, Work, Sense, ActuateC i t C t ACommunicate, Compute, AccessSearch for info, Find directionsExtend your perception of the worldT k t t d l tiTrack content and location

How would you configure it?In a stand alone mode I fi d k d iIn a fixed networked environmentIn ad-hoc networked environment

Should it be visible?How do you interface with it?How do you power it?

Content ExplosionNo shortage of content, either from private, corporate or public sourcesAggregation of content its structuring and indexing are key issuesExabyte (EB)

Content ExplosionAggregation of content, its structuring and indexing are key issuesFive exabytes of information is equivalent in size to the information contained in half a million new libraries the size of the Library of Congress print collections.

Exabyte (EB)1,000,000,000,000,000,000 bytes OR 10*18 bytes

2 Exabytes: Total volume of information generated in 1999.5 E b t All d k b h b i

% Change 1999-20001999-20002002 Terabytes2002 TerabytesStorage Medium

ScannedScanned CompressedCompressed

5 Exabytes: All words ever spoken by human beings.

Upper Estimates

Lower estimateUpper estimateLower estimateUpper estimate

36%2401,2003271,634Paper

-3%58,209431,69076,69420,254Film

80%2,073,7602,779,7603,416,2304,999,230Magnetic

69%69%2,132,2382,132,2383,212,7313,212,7313,416,2813,416,2815,421,2215,421,221TotalTotal

28%298151103Optical

36%2401,2003271,634Paper

Source: http://www.sims.berkeley.edu/research/projects/how-much-info-2003/printable_report.pdf Migrating to digital media

Wireless Evolution F Wireless Evolution Focus: UserUser--contentcontent

Focus: BandwidthBandwidth

Subscribers

Focus:

>QoE> Simplicity

> Performance

>Broadband

>New ServicesF

GrowthGrowth > Service Richness

>Security/trust

>Price

>EfficiencyFocus: CoverageCoverage

>Voice Quality >Scalability

>Ubiq it

>Coverage

>Mobility

>Portability

>Capacity

>Ubiquity

>Price

Voice

Mobile Ubiquitous NetworkingMobile Ubiquitous NetworkingUbiquitousUbiquitousMobile WorldMobile World

Broaden communication parties, networking, and business opportunities

Ubiquitous Ubiquitous WorldWorld

Networks with low performance devices(e.g. RF tags and sensors)

Mobile WorldMobile World

(Real World)

B3G Radio Access

B3G Mobile NetworkB3G Mobile Network

Networks with high performance Networks with high performance devices(e.g. home appliances)

Mobile NW Ubiquitous Local NW

Mobile EdgeMobile Edge

Mobile-Ubiquitous NW

Deep NetworkingSmall, inexpensive, lowSmall, inexpensive, low--powered sensors and actuators, deeply embedded into the physical powered sensors and actuators, deeply embedded into the physical

Deep Networking

environment in very large numbers, interacting and forming wired and wireless networks to environment in very large numbers, interacting and forming wired and wireless networks to communicate, adapt, act, respond, and coordinate highcommunicate, adapt, act, respond, and coordinate high--level tasks.level tasks. (e.g. Monitor the state of devices and provide necessary online diagnostics, maintenance support, etc.)

As these micro devices are networked the Internet will be pushed deeper than the current network As these micro devices are networked the Internet will be pushed deeper than the current network As these micro devices are networked, the Internet will be pushed deeper than the current network As these micro devices are networked, the Internet will be pushed deeper than the current network edge. This will enable a hundredfold increase in the size of the Internet.edge. This will enable a hundredfold increase in the size of the Internet.

Challenges:Challenges:New methods of networking devices to one another and to the Internet must New methods of networking devices to one another and to the Internet must b d l db d l dbe developed.be developed.

Smart dustSmart dust

D7024E MDCS: Introduction http://www-bsac.eecs.berkeley.edu/~warneke/SmartDust/index.html

Context aware devicesContext aware devices

Typical Mobile Computing ApplicationTypical Mobile Computing Application

DS: Challenges & design issues

TransparenciesTransparencies

•Access transparency: enables local and remote resources to be accessed using identicaloperations.•Location transparency: enables resources to be accessed without knowledge of their physical

t k l ti (f l hi h b ildi IP dd )or network location (for example, which building or IP address).•Concurrency transparency: enables several processes to operate concurrently using shared resources without interference between them.•Replication transparency: enables multiple instances of resources to be used to increase Replication transparency: enables multiple instances of resources to be used to increase reliability and performance without knowledge of the replicas by users or application programmers.•Failure transparency: enables the concealment of faults, allowing users and application programs to complete their tasks despite the failure of hardware or software components.•Mobility transparency: allows the movement of resources and clients within a system without affecting the operation of users or programs.P f t ll h b fi d i f •Performance transparency: allows the system to be reconfigured to improve performance as

loads vary.•Scaling transparency: allows the system and applications to expand in scale without change to the system structure or the application algorithms


the system structure or the application algorithms.

Architectural models & paradigms


Layered View on a Distributed System Layered View on a Distributed System

Applications & servicespp

DBMS, TPS, ...

Distributed OSMiddleware

m

Hardware Plat

form

Distributed Computing: System architecturesDistributed Computing: System architectures

The client/server modelPeer-to-peerpVariations to the 2 above

Figure 2.2Clients invoke individual servers

Client i ti Server

Server

Client invocation

result

Serverinvocation

resultServer

ClientClient

Process:Key:

Computer:


Figure 2.3A di ib d li i b d A distributed application based on peer processes

Peer 2

Application

ApplicationPeer 1

Application

Peer 3Sharableobjects

Peer 4

Peers 5 .... N

Application


Figure 2.4A service provided by multiple servers

Service

ServerServer

Client

Server

Client

Server


Figure 2.5Web proxy server

WebClient

Proxy

Web

server

server

server

Web Clientserver

Client

Caching


Figure 2.6Web applets) li t t lt i th d l di f l t da) client request results in the downloading of applet code

Web serverApplet code

Client

b) client interacts with the applet Mobile code

ClientWeb

AppletClient serverApplet


Figure 2.7Thin clients and compute servers

Network computer or PCCompute server

Thin Application

Network computer or PC

networkThinClient

ApplicationProcess

network


DS: Design IssuesDS: Design Issues

Must be transparentProvide flexibilityyHeterogeneityR li bilit /d d bilitReliability/dependabilityGood performanceSecurityScalableScalable

DS: ReliabilityDS: Reliability

Availability is a related factorDesign should not require the simultaneous g qfunctioning of a substantial number of critical componentspMore redundancy greater availability and greater inconsistencyinconsistencyFault tolerance, the ability to mask failures from the user

DS: Performance IssuesDS: Performance Issues

The rest are useless without thisHard to measure, benchmarks are meaningless, gBalance number of messages and grain size of distributed computationsdistributed computations

DS: ScalabilityDS: Scalability

A maxim for developing distributed systemsAvoid centralised components, tables and p ,algorithmsOnly decentralised algorithms should be usedOnly decentralised algorithms should be used

Scalability ProblemsScalability Problems

Figure 1-3. Examples of scalability limitations.


Scalability ProblemsScalability Problems

Characteristics of decentralized algorithms:No machine has complete information about the system state.Machines make decisions based only on local information.Failure of one machine does not ruin the algorithm.gThere is no implicit assumption that a global clock exists.


Pitfalls when Developing Distributed SystemsPitfalls when Developing Distributed Systems

False assumptions made by first time developer:The network is reliable.The network is secure.The network is homogeneous.gThe topology does not change.Latency is zeroLatency is zero.Bandwidth is infinite.Transport cost is zeroTransport cost is zero.There is one administrator.


Fundamental models


Fundamental modelsFundamental models

InteractionSynchronousAsynchronousEvent orderingg

FailureS itSecurity


Figure 2.8Real-time ordering of events

d i isend receive

1 4X

receive

m2

m1

receive

send2 3Y Physical

time

receivem2

Zreceive receive

send

m1 m2A

m3

receive receive

Areceive receive receive

t1 t2 t3


Figure 2.9Processes and channels

process p process q

send receivem

Communication channel

Outgoing message buffer Incoming message buffer


Figure 2.10Omission and arbitrary failures

Class of failure Affects DescriptionFail-stop Process Process halts and remains halted. Other processes may

detect this state.Crash Process Process halts and remains halted. Other processes may

not be able to detect this state.Omission Channel A message inserted in an outgoing message buffer never

arrives at the other end’s incoming message buffer.Send-omission Process A process completes a send, but the message is not put

in its outgoing message buffer.R i i i P A i i ’ i i Receive-omission Process A message is put in a process’s incoming message

buffer, but that process does not receive it.Arbitrary(B ti )

Process orh l

Process/channel exhibits arbitrary behaviour: it mayd/t it bit t bit ti(Byzantine) channel send/transmit arbitrary messages at arbitrary times,

commit omissions; a process may stop or take anincorrect step.


Figure 2.11Timing failures

Class of Failure Affects DescriptionClock Process Process’s local clock exceeds the bounds on its

rate of drift from real time.Performance Process Process exceeds the bounds on the interval

bbetween two steps.Performance Channel A message’s transmission takes longer than the

stated bound.


Figure 2.12Objects and principals

invocationObjectAccess rights

resultClient

Serverresult

NetworkPrincipal (user) Principal (server)


Figure 2.13The enemy

Copy of m

The enemym’

Communication channelProcess p Process qm


Figure 2.14Secure channels

P i i lA PrincipalBPrincipal A

S h lP P

Principal B

Secure channelProcess p Process q


DMCS: ConclusionsDMCS: Conclusions

ArchitecturesConcepts and terminologyp gyProblems of distributed & mobile computingD i i i lDesign principles

What’s next ? - Enabling technologiesWhat s next ? Enabling technologies

Radio-based communicationsWireless LANsEtc

d7024e: distributed & mobile computing systems

Documents