Distributed Systems 1-1 Outline What is a Distributed System? Examples of Distributed Systems Distributed System Requirements in Distributed System 1-2 What is a Distributed System? 1-3 1
What is a Distributed System? Component 1 Component n Component 1 Component n Middleware Middleware Network Operating System Network Operating System Hardware Hardware Host Host n-1 2 Component 1 Component n Middleware Network Operating System Hardware Host 1 Network Component 1 Component n Middleware Network Operating System Hardware Host n 1-4 What is a Distributed System? A distributed system is a collection of autonomous hosts that that are connected through a computer network. Each host executes components and operates a distribution middleware, which enables the components to coordinate their activities in such a way that users perceive the system as a single, integrated computing facility. 1-5 Middleware Examples Transaction-oriented IBM CICS BEA Tuxedo IBM Encina Microsoft Transaction Server Message-oriented Microsoft Message Queue NCR TopEnd Sun Tooltalk Procedural Sun ONC Linux RPCs OSF DCE Object-oriented OMG CORBA Sun Java/RMI Microsoft COM Sun Enterprise Java Beans 1-6 2
Centralised System Characteristics One component with non-autonomous parts Component shared by users all the time All resources accessible Software runs in a single process Single Point of control Single Point of failure 1-7 Distributed System Characteristics Multiple autonomous components Components are not shared by all users Resources may not be accessible Software runs in concurrent processes on different processors Multiple Points of control Multiple Points of failure 1-8 Examples of Distributed Systems 1-9 3
Motivation Case Studies Hongkong Telecom s Video-on-demand UBS IT Services Infrastructure Boeing s Aircraft Configuration Management Managing of Football Association Used the principles and techniques presented in this course. Will provide a motivation for this course. Will provide illustrative examples throughout this course... 1-10 Hongkong Telecom Video on demand Aim: provide subscribers with facilities to download videos from HK TK servers to low-cost Web-TVs. currently 90,000 users. Built using distributed object-technology. 1-11 Requirements Hardware: Clients: Web-TV Servers: RISC processor Operating System Heterogeneity : Clients: Java OS Servers: UNIX Programming Language Heterogeneity: Clients: Java Servers: C++ 1-12 4
Requirements (cont d) Communication across Network How to transmit complex data structures across the Internet? Scale Scaling from initially several hundred to currently 90,000 users Security Secure Payment Authentication 1-13 Why distributed object technology? Distributed: Video clients need to download/show video on customer s Web-TV Multiple servers needs to be operated by Hongkong Telecom: Object Technology: Video clients are written in Java: Web-TV has Java Virtual Machine portability to e.g. Sony Playstation, Sega-Console... Video servers are written in C++: high performance 1-14 IT Infrastructure of UBS Customer Information Services Authorisation Services Product Database Services Trading Workstation Marketing Services Host Services 1-15 5
Requirements Time to market Development of new applications with recent technology Integration of new applications increasingly difficult Scalability Management of 30,000,000 accounts Management of 10,000,000 customers Use by 2,000 concurrent users Reliability 1-16 Requirements (cont d) Hardware Heterogeneity Unisys Mainframes IBM Mainframes SPARC Servers PC Workstations Operating System Heterogeneity MVS UNIX Win-NT Programming Language Heterogeneity Cobol C/C++ Visual Basic 1-17 Why distributed object technology? Uniform view of all banking services Appropriate level of abstraction Preserving investment by wrapping legacy applications Exploiting advantages of object technology for new development Resolving distribution heterogeneity 1-18 6
Boeing 777 Configuration Mgmnt. 1-19 Problems to be solved Scale 3,000,000 parts per aircraft Configuration of every aircraft is different CAA regulations demand that records are kept for every single part of aircraft Aircraft evolve during maintenance Boeing produce 500 aircraft per year Configuration database grows by 1.5 billion parts each year Projected life of each aircraft 30 years 45,000 engineers need on-line access to engineering data 1-20 Problems to be solved (cont d) COTS Integration Existing IT infrastructure was no longer appropriate Boeing could not afford to build required IT infrastructure from scratch Components were purchased from several different specialized vendors relational database technology enterprise resource planning computer aided project planning Components needed to be integrated 1-21 7
Problems to be solved (cont d) Heterogeneity 20 Sequent database machines as servers for the engineering data 200 UNIX application servers NT and UNIX workstations for engineers 1-22 Why distributed object technology Object wrapping of COTS Resolution of distribution at high level of abstraction Resolution of heterogeneity Scalability 1-23 Management of Football Association Managing soccer leagues, national team, clubs, player transfer Imaginary system Common example that can be twisted for didactic purposes 1-24 8
Requirements Autonomy of clubs Every club operates its own administration, training/game scheduling. Need for integration in order to register players with the football association. book players for national team games. agree to schedule of league games. Heterogeneity different machines different programming languages 1-25 Distributed System Requirements 1-26 Requirements Integration of new, legacy and components off-the-shelf Legacy components might not need to be reengineered COTS cannot be modified Heterogeneity of hardware platforms operating systems networks programming languages Construction of distributed systems 1-27 9
Common Requirements What are we trying to achieve when we construct a distributed system? Certain requirements are common to many distributed systems Resource Sharing Openness Concurrency Scalability Fault Tolerance 1-28 Resource Sharing Ability to use any hardware, software or data anywhere in the system. Resource manager controls access, provides naming scheme and controls concurrency. Resource sharing model (e.g. client/ server or object-based) describing how resources are provided, they are used and provider and user interact with each other. 1-29 Openness Openness is concerned with extensions and improvements of distributed systems. Detailed interfaces of components need to be published. New components have to be integrated with existing components. Differences in data representation of interface types on different processors (of different vendors) have to be resolved. 1-30 10
Concurrency Components in distributed systems are executed in concurrent processes. Components access and update shared resources (e.g. variables, databases, device drivers). Integrity of the system may be violated if concurrent updates are not coordinated. Lost updates Inconsistent analysis 1-31 Scalability Adaption of distributed systems to accomodate more users respond faster (this is the hard one) Usually done by adding more and/or faster processors. Components should not need to be changed when scale of a system increases. Design components to be scalable! 1-32 Fault Tolerance Hardware, software and networks fail! Distributed systems must maintain availability even at low levels of hardware/software/network reliability. Fault tolerance is achieved by recovery redundancy 1-33 11
in Distributed Systems 1-34 Distributed systems should be perceived by users and application programmers as a whole rather than as a collection of cooperating components. has different dimensions that were identified by ANSA. These represent various properties that distributed systems should have. 1-35 Distribution Scalability Performance Failure Migration Replication Concurrency Access Location 1-36 12
Access Enables local and remote information objects to be accessed using identical operations. Example: File system operations in NFS. Example: Navigation in the Web. Example: SQL Queries 1-37 Location Enables information objects to be accessed without knowledge of their location. Example: File system operations in NFS Example: Pages in the Web Example: Tables in distributed databases 1-38 Concurrency Enables serveral processes to operate concurrently using shared information objects without interference between them. Example: NFS Example: Automatic teller machine network Example: Database management system 1-39 13
Replication Enables multiple instances of information objects to be used to increase reliability and performance without knowledge of the replicas by users or application programs Example: Distributed DBMS Example: Mirroring Web Pages. 1-40 Failure Enables the concealment of faults Allows users and applications to complete their tasks despite the failure of other components. Example: Database Management System 1-41 Migration Allows the movement of information objects within a system without affecting the operations of users or application programs Example: NFS Example: Web Pages 1-42 14
Performance Allows the system to be reconfigured to improve performance as loads vary. Example: Distributed make. 1-43 Scaling Allows the system and applications to expand in scale without change to the system structure or the application algortithms. Example: World-Wide-Web Example: Distributed Database 1-44 Key Points What is a Distributed Systems Adoption of Distributed Systems is driven by Non-Functional Requirements Distribution needs to be transparent to users and application designers has several dimensions dimensions depend on each other 1-45 15