Ds1 int (1)
- 2. 2 Course/Slides Credits Note: all course presentations are based on those developed by Andrew S. Tanenbaum and Maarten van Steen. They accompany their "Distributed Systems: Principles and Paradigms" textbook (1st & 2nd editions). http://www.prenhall.com/divisions/esm/app/aut hor_tanenbaum/custom/dist_sys_1e/index.html And additions made by Paul Barry in course CW046-4: Distributed Systems http://glasnost.itcarlow.ie/~barryp/net4.html
- 3. 3 Definition of a Distributed System (1) A distributed system is: A collection of independent computers that appears to its users as a single coherent system
- 4. 4 Definition of a Distributed System (2) A distributed system organized as middleware. The middleware layer extends over multiple machines, and offers each application the same interface
- 5. 5 Goals of Distributed Systems • Easily Connect Users/Resources • Exhibit Distribution Transparency • Support Openness • Be Scalable: – in size – geographically – administratively Looking at these goals helps use answer the question: “Is building a distributed system worth the effort?”
- 6. 6 Transparency in a Distributed System Different forms of transparency in a distributed system (ISO, 1995)
- 7. 7 Scalability Limitations Examples of scalability limitations Concept Example Centralized services A single server for all users Centralized data A single on-line telephone book Centralized algorithms Doing routing based on complete information
- 8. 8 Scaling Techniques (1) 1.4 The difference between letting (a) a server or (b) a client check forms as they are being filled
- 9. 9 Scaling Techniques (2) 1.5 An example of dividing the DNS name space into zones
- 10. 10 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. • There is no implicit assumption that a global clock exists.
- 11. 11 Pitfalls when Developing Distributed Systems • The network is reliable • The network is secure • The network is homogeneous • The topology does not change • Latency is zero • Bandwidth is infinite • Transport cost is zero • There is one administrator
- 12. 12 Types of Distributed Systems • Distributed Computing Systems – High Performance Computing (HPC) • Distributed Information Systems – Transaction Processing Systems (TPS) – Enterprise Application Integration (EAI) • Distributed Pervasive Systems – Ubiquitous Systems
- 13. 13 A. Frank - P. Weisberg Clustered Systems Architecture
- 14. 14 Cluster Computing Systems • Collection of similar workstations/PCs, closely connected by means of a high-speed LAN: – Each node runs the same OS. – Homogeneous environment – Can serve as a supercomputer – Excellent for parallel programming • Examples: Linux-based Beowulf clusters, MOSIX (from Hebrew University).
- 15. 15 Architecture for Cluster Computing System
- 17. 17 Grid Computing Systems • Collection of computer resources, usually owned by multiple parties and in multiple locations, connected together such that users can share access to their combined power: – Can easily span a wide-area network – Heterogeneous environment – Crosses administrative/geographic boundaries – Supports Virtual Organizations (VOs) – Examples: EGEE - Enabling Grids for E-SciencE (Europe), Open Science Grid (USA).
- 18. 18 Architecture for Grid Computing Systems
- 19. 19 Cloud Computing Systems (1) • Collection of computer resources, usually owned by a single entity, connected together such that users can lease access to a share of their combined power: – Location independence: the user can access the desired service from anywhere in the world, using any device with any (supported) system. – Cost-effectiveness: the whole infrastructure is owned by the provider and requires no capital outlay by the user. – Reliability: enhanced by way of multiple redundant sites, though outages can occur, leaving users unable to remedy the situation.
- 20. 20 Cloud Computing Systems (2) – Scalability: user needs can be tailored to available resources as demand dictates – cost benefit is obvious. – Security: low risk of data loss thanks to centralization, though problems with control over sensitive data need to be solved. – Readily consumable: the user usually does not need to do much deployment or customization, as the provided services are easy to adopt and ready-to-use. • Examples: Amazon EC2 (Elastic Compute Cloud), Google App Engine, IBM Enterprise Data Center, MS Windows Azure, SUN Cloud Computing.
- 21. 21 Transaction Processing Systems (TPS) The role of a TP monitor in distributed systems
- 23. 23 Communication Middleware Models/Paradigm • Distributed File Systems • Remote Procedure Call (RPC) • Distributed Objects (RMI) • Distributed Documents
- 24. 24 Distributed Pervasive Systems • Requirements for pervasive systems: – Embrace contextual changes – Encourage ad hoc composition – Recognize sharing as the default – Support distribution transparency
- 25. 25 Electronic Health Care Systems (1) • Questions to be addressed for health care systems: – Where and how should monitored data be stored? – How can we prevent loss of crucial data? – What infrastructure is needed to generate and propagate alerts? – How can physicians provide online feedback? – How can extreme robustness of the monitoring system be realized? – What are the security issues and how can the proper policies be enforced?
- 26. 26 Electronic Health Care Systems (2) Monitoring a person in a pervasive electronic health care system, using (a) a local hub or (b) a continuous wireless connection
- 27. 27 Sensor Networks (1) • The nodes to which sensors are attached are: – Many (10s-1000s). – Simple (i.e., hardly any memory, CPU power, or communication facilities). – Often battery-powered (or even battery-less). • Questions concerning sensor networks: – How do we (dynamically) set up an efficient tree in a sensor network? – How does aggregation of results take place? Can it be controlled? – What happens when network links fail?
- 28. 28 Sensor Networks (2) Organizing a sensor network database, while storing and processing data (a) only at the operator’s site or …
- 29. 29 Sensor Networks (3) Organizing a sensor network database, while storing and processing data … or (b) only at the sensors