Distributed Systems

Distributed Systems (218 - 15754)
Study: Bachelor in Informatics Engineering
Semester 2/Spring Semester
3rd Year Course/Upper Division

Students are expected to have completed:

Operating Systems
Computer networks

Compentences and Skills that will be Acquired and Learning Results:

The main objective of the course is to describe the main concepts needed for designing and developing distributed systems and applications.

1. Generic competences:
- Capacity of analysis and synthesis (PO a, b, c, e)
- Capacity of organization and scheduling (PO c, d)
- Problem solving (PO a, b, c, g)
- Capacity to apply theoretical concepts (PO a, c)

2. Specific competences:
a. Cognitive (PO a,c e,j)
- Concepts of a distributed system
- Main hardware platforms used in distributed systems
- Understanding the concept of middleware and the services provided by a middleware
- Techniques for developing distributed applications
- Storage systems used in distributed systems
- Techniques for developing fault tolerant distributed systems
- Knowledge and application of features, functionality and structure of Distributed Systems, Computer Networks and Internet-based design and implement distributed applications (CECRI11)
- Ability to design and implement software and communication systems (CEIC4)

b. Procedimental/Instrumental (PO a, b, c, g, j, k)
- Designing and developing a distributed system and a distributed application.
- Designing and developing distributed applications using the main TCP/IP services, RPC or services provided by middlewares.
- Using tools (programming languages and operating systems) appropriates for developing distributed systems.
- Designing and developing distributed applications using middleware services and remote procedure calls.
- Evaluating the performance of a distributed application
- Researching and looking for solutions to new problems related to distributed systems.

c. Attitude (PO: c, e)
- Creativity
- Critical vision of current distributed systems
- Motivation
- Interest for acquiring new knowledge and information

Description of Contents: Course Description

The basic concepts of this course are: concurrency; interprocess communication; middleware; RPC; distributed file systems; distributed applications; fault tolerance; web services;

Content:

1. Introduction
- Basic concepts
- Interconnection networks
- Advantages of distributed systems
- Distributed computing paradigms
- Design of distributed systems

2. Process communication and synchronization
- Communication mechanisms in shared memory systems
- Communication mechanisms in distributed memory systems
- POSIX services
- Threads in Java

3. Message passing and client-server applications
- Communication model using message passing
- Design aspects
- POSIX queue messages
- Client-server applications
- Design of concurrent servers

4. Communication using sockets
- Communication model with sockets
- POSIX sockets API
- Java sockets API
- Design guide of client-servers applications using sockets

5. Remote procedure call
- RPC behavior
- Interface definition language
- Marshaling and message transfer
- ONC-RPC
-Remote method invocation in JAVA (RMI)

6. Web services
- HTTP protocol
- SOAP
- Development of web services with JAX-WS
- Development of web services with gSOAP

7. Distributed services
- Name services
- Synchronization in distributed systems
- Physical and logical clocks
- Distributed mutual exclusion
- Algorithms of election
- Multicast

8. Distributed storage systems
- Distributed file systems structure
- File and directory services
- Implementation of distributed file systems
- Example: NFS
- Shared disks file systems
- Parallel file systems
- Storage area networks

9. Fault tolerant in distributed systems
- Fault tolerant concepts
- Software fault tolerance
- Fault detectors
- Replication
- Protocols of consensus

Learning Activities and Methodology

The teaching methodology includes:
1. Theoretical lectures 1 ECTS (PO a, b, c, e, j)
- Distributed systems concepts
- Basic text books will be also recommended
2. Projects 1.5 ECTS (PO a, b, c, d, e, g, k)
- Several projects are made along the course applying the concepts shown in theoretical lectures. Partial teacher support in computer labs.
- Students have to analyze the requirements and provide a working solution
- The projects are developed in team of 2 students, in order to promote the collaborative work
3. Problem solving with the teacher 1 ECTS (PO a, c, e)
- By solving exercises and case of studies in a participatory way.
4. Student work 2 ECTS. (PO a, c, e, g, k)
- Self-study to understand the theoretical concepts
- Homework for solving proposed exercises
- Students must to understand and present in public the main aspects obtained in the analysis of a basic paper written in English.
5. Exams 0.5 ECTS (PO a, c, e, g)
- Midterm and final exam

Assessment System:

The evaluation includes the following parts:

1. The continuous assessment (70 %) includes:
- Programming and laboratory projects: 30%. (PO b, c, d, e, g, k)
- Analyze and public presentation of a simple and basic paper written in English: 15% (b, g, j)
- Problem solving: 25% (PO a, c, e, g, j). Students will have to do several exercises.

2. The percentage of the final exam is: 30%. (PO a, c, e, g, j). The final exam will include theoretical and practical concepts.
The minimum value for the final exam will be 4.
The minimum value for all lab projects will be 4.

A student follows the continuous assessment when the student:

- makes all labs, the minimum value for each lab will be 2.
- the mínimum values for all lab projects is 4.
- makes the article presentation.

The final exam in the extraordinary period will include the theoretical and practical concepts of the course.

The final grade will be increased by 1 point to those students who complete all parts of the continuous assessment, obtain more than 7 in the continuous evaluation, and at least 4 in the final exam.

Basic Bibliography:

G. Coulouris, J. Dollimore, T. Kindberg, G. Blair. Distributed Systems, Concepts and design. 5ª edition. 2011. Addison-Wesley.

Additional Bibliography:

Distributed Systems: principles and paradigms. Andrew S. Tanenbaum , Maarten van Steen. Pearson.
F. García, J. Carretero, A. Calderón, J. Fernández, J. M. Pérez. Problemas resueltos de programación en C. Thomson.
Pankaj Jalote. Fault Tolerance in Distributed Systems. Prentice-Hall.
Richard Stevens. UNIX Network Programming. Prentice Hall.