Programming Distributed Systems

Organization

Instructor: Dr. Annette Bieniusa

Lecture: Monday, 10:00 - 11:30, Room 48-453

Lab session: Wednesday, 15:30 - 16:30, Room 32-411 (in the lab session we will support you, if you have problems solving the practical exercises; no new material will be covered here)

Exercises: Thursday, 15:30 - 17:00, Room 42-110 (first session on April 12th)

The lecture and course material will be in English.

Registration:

Please register via mail to Peter Zeller (subject: “Registration Progdist”).

Prerequisites:

  • Good programming knowledge, including usage of code repositories (git)

As this is a 4CP lecture, you will spend on average around 120 hours on the lecture and the exercises, plus exam preparation. If you do not have much experience with programming systems of medium size, you need to plan to spend more time on the practical exercises! If you lack some of the theoretical prerequisites (definition of formal languages, finite-state machines), you should also incorporate this into your planning.

Objectives

You will be able to

  • explain the challenges regarding time and faults in a distributed system
  • provide formal definitions for time models, fault models and consistency models
  • comprehend and develop models of some distributed system in a process calculus
  • describe the algorithms for essential abstractions in programming distributed systems
  • implement basic abstractions for distributed programming
  • explain the virtues and limitations of major distributed programming paradigms

Topics of the Lecture

  • Basic primitives in programming distributed systems
    • Broadcast
    • Consistent snapshots
    • Consensus
    • Distributed state machine replication
  • Theoretical foundation
    • Failure modes
    • Time in Distributed systems
    • Process calculi (e.g. CSP, CCS, pi calculus, TLA)
    • Consistency models
  • Programming paradigms for Distributed Systems
    • RPC
    • Futures and promises
    • Message passing and Actors
    • CRDTs
    • Streaming / Data-Flow

Exercises

To solve the practical exercises, you need to install some software: Required software for the exercises

Please submit your solution to the practical exercises via your Git repository (which you get after registration). If you want feedback on the theoretical exercises, please send your solution to Peter Zeller.

Solving the exercise sheets is not mandatory, but highly recommended. In the second half of the semester, you will have to implement a bigger project in Erlang. Successfully implementing this project is a requirement for being admitted to the exams.

Sheet Material
Exercise 1 ex1_tests.erl
Exercise 2 template
Date Topic
12.04. Introduction to Erlang (Slides)
19.04. Discussing sheet 1, introduction to concurrency in Erlang (Slides)

Time Table

Week Date Topic
1 09.04. Introduction
2 16.04. Message Broadcast
3 23.04. Implementing Broadcast
4 30.04. Replication
5 07.05. Consensus I (2PC/3PC)
6 14.05. Consensus II (Paxos, FLP)
7 21.05. — (Pentecost)
8 28.05. Consistency I
9 04.06. Consistency II
10 11.06. Conflict-Free Replicated Datatypes
11 18.06. Verification and Testing
12 25.06. Programming models: RPC, Actors
13 02.07. Programming models: Streams, Parallel Batch Processing
14 09.07. Conclusion

Please report errors and give feedback via our issue tracker or by mail.