OVERVIEW OF DISTRIBUTED MULTIMEDIA COLLABORATION

Tao Huang

01-09-2008

OUTLINE

Distributed Systems Multimedia Collaboration HCI design in Collaboration Architecture Design of Collaboration Systems

2

DISTRIBUTED SYSTEMS – DEFINITION Definition

A collection of (probably heterogeneous) automata whose distribution is transparent to the user so that the system appears as one local machine.

Automata coordinate their activities through a distribution middleware

Classification DFS: Distributed File System DOS: Distributed Operation System

Pros & Cons Transparent to network Resource sharing Fault tolerant Based on flawed assumptions

3

DISTRIBUTED SYSTEMS – HISTORY

1970-1985 IBM PC, RFC 791(IP), Ethernet, ARPANET

1985-1995 Sun NFS, RFC 1510(Kerberos V5), GPRS

1995-Current Internet/World Wide Web Gigabit Ethernet, Fiber optics P2P, Napster Soap, Web Services, Grids

4

DISTRIBUTED SYSTEMS – ARCHITECTURES

Typical architectures Client Server:

Smart client fetch and display data from server 3-tier:

Move client intelligence to middle tier N-tier:

Forward requests to other enterprise services Clustered:

Tasks are subdivided and processed on integrated machines,

results will be put together

5

DISTRIBUTED SYSTEMS – ARCHITECTURES II

Typical architectures Parallel Computing

Simultaneous use of multiple compute resources to solve a computational problem

Distributed Objects RM-ODP & CORBA

Peer-to-Peer SOA(Service Oriented Architecture)

Web services Grid services Web 2.0

6

DISTRIBUTED SYSTEMS – STANDARDS & TECHNOLOGIES

Related standards RM-ODP (ISO/ITU-T) OMG CORBA Open Group’s DCE W3C Web Service Architecture (WS-*)

Related platforms Microsoft .Net Framework Java VM ADAPTIVE Communication Environment (ACE)

Related techniques Java RMI SOAP 7

DISTRIBUTED SYSTEMS – MIDDLEWARE Richard E. Schantz et al.

“Middleware for Distributed Systems”

8

DISTRIBUTED SYSTEMS – MIDDLEWARE CLASSIFICATION

Host infrastructure middleware Encapsulates and enhances native OS

communication and concurrency mechanisms Distribution middleware

Defines higher-level distributed programming models

Common middleware services Augment distribution middleware by defining

higher-level domain-independent services Domain-specific middleware services

Tailored to the requirements of particular domains, such as telecom, e-commerce, health care, process automation, or aerospace

9

DISTRIBUTED systems – SOME RESEARCH CHALLENGES

From “Report of the NSF Workshop on Research Challenges in Distributed Computer Systems”(2005) Security

Making networks secure, Confidence in the environment Simplifying Management

Simplify management of technology across the various domains

Continuous operation/Fault tolerant Avoiding correlated failures, Living with failure. Failure

detection, Self configuration, Software upgrades, Better tools, Availability versus Consistency

Real time QoS Response time, jitter, throughput, availability, survivability,

recovery, time after attack/failure, call drop rate, access failure rate, packet delay, packet drop rate

10

Distributed Systems Multimedia Collaboration HCI design in Collaboration Architecture Design of Collaboration Systems

11

MULTIMEDIA COLLABORATION – DEFINITION Multimedia

Handling a variety of representation media in an integrated manner

It has both continuous and discrete media types Collaboration

Structured, recursive process where two or more people work together toward a common goal

Shared generic tools: Text chat, White boards, Audio/Video conferencing

Shared applications: Web pages, PowerPoint Some Typical Multimedia Collaborations

Audio/Video Conferencing E-Learning Online Social Communities

12

MULTIMEDIA COLLABORATION – EXAMPLE SYSTEMS Sync Collaboration

• Access Grid & Virtual Room Video conferencing System(VRVS)

• Admire from BUAA• Polycom• Skype• Gtalk• WebEx• GlobalMMCS from IUB

Async Collaboration• E-Learning

• Sakai Project (e.g. IU Oncourse), Blackboard

• Online Social Communities• FaceBook, MySpace & Second Life

13

MULTIMEDIA COLLABORATION – SOME CHALLENGES OF SYNC MULTIMEDIA

Supporting streaming multimedia Programming model support System support

Quality of service support QoS classification

Timeliness dimensions Volume dimensions Reliability dimensions

Static & dynamic QoS management Real time synchronization support

intra- & inter-media synchronization Multiparty communications 14

MULTIMEDIA COLLABORATION – SYSTEM I PARTICIPATED IN Global-MMCS

A service-oriented collaboration system, which  integrates various services including videoconference, instant messaging and streaming

Web services based conference control framework

Based on XGSP (XML based General Session Protocol)

Integrates heterogeneous systems into one collaboration system

15

MULTIMEDIA COLLABORATION – MY CURRENT WORK

CGL Annotation System

16

MULTIMEDIA COLLABORATION – MY CURRENT WORK

17

Distributed Systems Multimedia Collaboration HCI design in Collaboration Architecture Design of Collaboration Systems

18

HCI DESIGN IN COLLABORATION – DEFINITION

Interaction Design(ID) “Designing interactive products to support the way

people communicate and interact in their everyday and working lives”

The center of ID is to achieve good user experience Human Computer Interaction(HCI), a subset

of ID Being “concerned with the design, evaluation, and

implementation of interactive computing systems…”(ACM SIGCHI, 1992, p.6)

People involved in ID Designers/Developers Users/Stakeholders

19

HCI DESIGN IN COLLABORATION – DESIGN PROCEDURE

The procedure of ID Identify needs and establish requirements for user

experience Develop alternate designs Building interactive versions of the designs Evaluation

Evaluation of ID Usability goals

Effectiveness, Efficiency, Safety, Utility, Learnability, Memorability

User experience goals Design principles

Visibility, Feedback, Constraints, Consistency, Affordance

20

HCI DESIGN IN COLLABORATION – IMPORTANT DESIGN ISSUES

Design in Collaboration Conversation Support

Communicating in physically different locations Communicating in co-located settings Computer mediated communication

Coordination Support Verbal/Non verbal communication Shared external representations

Awareness Support Overhearing and monitoring

Personalization Support GUI customization Status setup 21

Distributed Systems Multimedia Collaboration HCI design in Collaboration Architecture Design of Collaboration

Systems

22

ARCHITECTURE DESIGN OF COLLABORATION SYSTEMS – TOPOLOGY OF EXISTING SYSTEMS

Centralized architecture AG, VRVS and Admire: MBONE based, capable of

supporting other collaboration systems such as H.323

Windows Netmeeting and Polycom: highly relies on MCU

Windows Live, iChat AV and AOL IM Decentralized architecture

Skype, P2P overlay Hierarchical Hybrid

23

ARCHITECTURE DESIGN OF COLLABORATION SYSTEMS – TOPOLOGY CHARACTERISTICS

Centralized architecture Pro: Simple to monitor, control and secure Con: Difficult or expensive to scale for heavy load

Decentralized architecture Pro: Easy to scale the system to support heavy

load, fault-tolerant Con: Insecure and hard to keep coherent

Hierarchical Pro: Scalability, partially fault-tolerant Con: Root corruption may cause failure, difficult to

keep coherent

24

Questions ?

25

26

WHY DISTRIBUTED SYSTEMS?

Every application is part of your business model must make them work together!

Shipping/ Receiving

Inventory

Engineering

Manufacturing

Accounting

Payables/ Receivables

Sales

27

WHY DISTRIBUTED SYSTEMS?

Application Integration and Distributed Processing are the same thing Constructing information-sharing distributed systems from diverse

sources: heterogeneous networked physically disparate multi-vendor

27

28

WHY DISTRIBUTED SYSTEM? Application Requirements

Functional Non-Functional

Non-functional requirements drive distribution of a system Scalability Concurrency Openness Heterogeneity Resource sharing Fault-tolerance …

ADVANTAGES AND DISADVANTAGES OF DISTRIBUTED SYSTEMS

Advantages Sharability Expandability Local autonomy Improved performance Improved reliability and availability Potential cost reductions

Disadvantages Network reliance Complexities Security Multiple point of failure 29

Ahm

et F

atih

Musta

coglu

30

WEB SERVICES

Allows applications to communicate with each other in a platform- and programming language-independent manner

Operations in a Web Service Architecture: Publish, Find and Bind

Based upon the interactions between three roles: service provider, service registry and service requestor

Web Services interact by exchanging messages in SOAP format

De facto specifications: SOAP WSDL UDDI

SOAP Stands for Simple Object Access Protocol Based on XML

Simple and extensible An envelope which includes

Header Body

A communication protocol Provide communication between applications

A format for sending messages Platform independent Language independent

31

COBLITZ I P2P style but based on existing http CDN Divide large files into large chunks instead of small

blocks An agent splits the large file request into requests of

multiple chunks, and it is responsible of merging fetched chunks

Proxy caches chunks for later client requests

32

COBLITZ II

Coping With Scale Unilateral, Asynchronous Peering

Make peering a unilateral, asynchronous decision Peer Set Selection

Use minimum application-level ping times when determining suitable peers

Apply hysteresis to the peer set

Request Routing Highest Random Weight

33

DESIGN PRINCIPLES – VISIBILITY & CONSTRAINTS (OLD MRAS)

DESIGN PRINCIPLES – VISIBILITY & CONSTRAINTS (NEW MRAS)