Distributed Systems: Course Information

• Distributed Systems Class Information for the 3rd MIEIC is being taught by Pedro F. Souto ([email protected]) and the date of the class information is February 13, 2019

Staff

• The lectures are given by Pedro Ferreira do Souto ([email protected])
• The labs are run by different people: Hélder Fernandes Castro ([email protected]), Pedro Alves Nogueira ([email protected]), and Pedro Ferreira do Souto ([email protected])

Distributed Applications

• A distributed application has two or more processes
• These processes execute on different computers
• They communicate via messages
• There is a no negligible delay concerning computation
• Most, if not all, intellectually interesting applications are distributed nowadays

Course Objectives

• Understand the foundations of distributed computing
• Design and implement simple distributed applications
• Analyze distributed solutions and evaluate their fitness of solutions to the problem
• The course will focus on cloud-based computing instead of the Internet-of-Things

Prerequisites

• Requires knowledge of operating systems, specifically concurrency
• Requires knowledge of computer networks
• Requires Programming knowledge in C/C++/Java

Syllabus Parts

• Part 1: Communication and Processing which includes Introduction, Networking (Review), Communication Paradigms (Messages, Remote invocation), and Processing
• Part 2: Foundations includes Names and Localization, Security, Sinchronization, Replication and Consistency, and Fault Tolerance
• Part 3: Applications includes Distributed Filesystems, Data Intensive Computing, Web-based Applications and Peer-to-peer Systems

Supporting Material

• Tanenbaum, A. e van Steen, M.’s “Distributed Systems: Principles and Paradigms, 3rd Ed. (2017)” textbook is available for free upon request
• "Distributed Systems: for fun and profit" is recommended as a small fun book focused on the data center.
• Java documentation is also recommended, specifically Java 8.0 API

Mini-Projects

• Project 1 is a simple distributed storage application
• Project 1 should be done in groups of 2 students
• Project 1 is due on April 7 @ 20:00 (Sunday)
• Project 1 demo is on first lab class after due date (i.e. on April 8)
• Project 2 is one's own project
• Project 2 must use non-blocking API
• Project 2 should be done in groups of 4 students
• Project's 2 proposal is due April 14 @ 20:00 (Sunday)
• Project 2 is due on May 26 @ 20:00 (Sunday)
• Project 2's demo on first lab class after due date (i.e. last week)
• Both projects must be implemented in Java
• Both projects and their reports must be submitted via SVN/Redmine
• Both projects have the same weight

Mini-Projects Grading

• Grading is individual
• Each project is graded assuming the expected number of group members (2 and 4 respectively)
• A contribution factor computed from the contribution using a piecewise linear function is applied to the grade
• First project "Breaking points": 33%, 50%
• The factor for the first project is: 0 at 0%,.85 at 33%, 1 at 50% and 1.10 at 100%
• Second project "Breaking point": 25%
• The factor for the second project is: 0 at 0%, 1 at 25%, 1.15 at 100%
• A ceiling based on the complexity will also be applied
• Students can fill out a form to keep their previous year’s project grade before February 17th

Exam

• The exam is in Moodle
• The exam may include True/False, multiple-choice, and open questions
• The exam is closed books but allows a cheat sheet
• The cheat sheet must be A4 (both sides) and handwritten
• Exams that do not comply will be nullified
• You should study by the book

Class Participation

• Participation includes discussion in the Moodle's forum

Final Grade

• The formula to calculate the final grade (G) for ordinary students is: min(min(F, P) + 3, 0.45P + 0.1C + 0.45F)
• The parameters are:
  • P is the average of the grades in both projects (P ≥ 8)
  • C is class participation
  • F is the final exam grade (F ≥ 8)
• The final grade cannot exceed in more than 3 points (in 20) the minimum of the grades in the final exam and in the project
• The formula to calculate the final grade (G) for special status students is: min(min(F, P) + 3, 0.5P + 0.5F)

Academic Integrity

• Academic integrity is taken very seriously
• Students should check out the Declaração de Princípios sobre a Integridade Académica na UP
• Students are allowed to discuss the projects
• There will be a discussion forum on Moodle for each project
• The code submitted must be either developed by the group members or authorized, and credit given in both the report and the source file
• Tools will be used to automatically detect common code and all groups with similar code will be penalized
• You may still help your colleagues, but code cannot be shared nor can you read someone else’s project
• Projects that are identical to those of last year is not allowed
• The penalty for academic dishonesty/integrity ranges from a zero in that project to failing the course

Important Dates:

• Classes start 10 minutes after the hour
• Wednesdays at 11:10
• Fridays at 8:40
• Labs start next week
• Course material is available on Moodle’s course page
• 1st project is on April 7 @ 20:00 (Sunday)
• Proposal of 2nd project is on April 14 @ 20:00 (Sunday)
• 2nd project is on May 26 @ 20:00 (Sunday)
• Demos are in the first lab class after due dates.