Large-Scale Distributed Systems

This course presents the paradigms, algorithmic techniques and design principles for large-scale distributed systems (LSDS) design. The focus of the course is on decentralized, autonomous and self-organizing systems. It favors practical case studies of state-of-the-art system designs from academia and industry.

The course covers: LSDS characteristics and requirements, gossip-based protocols, DHTs and other routing/querying substrates, distributed file systems and distributed data management, large-scale processing & MapReduce, multicast, publish-subscribe, and content-based routing.

A unique feature for a course on LSDS is the use of hands-on experimentation of a set of representative algorithms on a distributed testbed using real protocols deployments and no simulation. This task is made simple thanks to a specialized framework developed at the university and used by several research groups worldwide.

Evaluation: small projects reports, and final exam based in part on an industry system example.
 

Learning Outcomes: 

On successful completion of this course, you will:

  • understand the key paradigms and services that one can use to build scalable applications;
  • understand the objectives of fault tolerance, load balancing and efficiency and what techniques are appropriate to achieve them;
  • have a general knowledge of the existing solutions used at major companies such as Amazon, Yahoo! or Google, and understand why the traditional solutions are not effective anymore;
  • have a practical experience in writing and evaluating code on a real large-scale testbeds, processing, presenting and discussing the results of your observation.
     
Type: 
Course
Semester: 
A2015
ECTS: 
5
Site: 
N
Code: 
12031
Language: 
english
Period: 
weekly
Schedule: 
Thursday: 14:00 - 18:00
Location: 
UniNE, Unimail
Room: 
B013
Evaluation type: 
written exam
Comment: 

First Lecture
The first lecture will take place on Thursday, 17.09.2015 at 14:00 in UniNE, Unimail, room B013.


ILIAS
The course page in ILIAS can be found at https://ilias.unibe.ch/goto_ilias3_unibe_crs_849824.html.