University Homepage School Homepage School Contacts School Search

Extreme Computing (2012/2013)

• Teaching Staff
• Time and Place
• Assessment
• Lectures
• Lab Sessions
• Assignments
• Reading List
• Previous guest lectures

Course Overview

Extreme Computing deals with cloud computing and cluster computing. It touches upons areas such as:

• [Background:] Motivation for new computing paradigms; introduction and differences between cloud and cluster computing; scaling, Big Data, performance, privacy, economics, security, software as service.
• [Infrastructure:] Distributed file systems; virtualisation; replication; fault tolerance; concurrent programming; web services.
• [Algorithms:] Hadoop (MapReduce); design and implementation of MapReduce programs; dealing with massive amounts of data; case studies using natural language processing, database query evaluation and/or machine learning. Streaming algorithms, randomised algorithms.

We aim to make students aware of topics using large (> 1000 node) clusters, but not be tied to particular vendors whereever possible. If you want to understand how companies like Google, Facebook, Twitter or Last.fm work, then this is the course for you.

Students should have a computing or numerate background, but there is no need to know any particular programming language. Students should be familiar with Unix (Linux) and ssh.

The course is suitable for all MSc students in Informatics. Final year undergraduates can take it with permission.

Guest Lecture: John Wilkes (Google)

This year we are delighted that John we give a guest lecture on Monday 24th September 2012 4-5pm:

Cluster Management at Google

Cluster management is the term that Google uses to describe how we control the computing infrastructure in our datacenters that supports almost all of our external services. It includes allocating resources to different applications on our fleet of computers, looking after software installations, and monitoring. My goal is to present an overview of these systems, introduce the new cluster-manager tool we are building, and present some of the challenges that we're facing along the way. Many of these challenges represent research opportunities, too, so I'll spend the majority of the time discussing those.

Bio

John Wilkes has been at Google since 2008, where he is working on cluster management and service level agreements for infrastructure services. He is interested in far too many aspects of distributed systems, but a recurring theme has been technologies that allow systems to manage themselves. In his spare time he continues, stubbornly, trying to learn how to blow glass.

Forum

The course has a forum which can be used to discuss the material of the course. Note that you will need an EASE account to log on.

Mailing List

Announcements regarding the course will be posted to the course mailing list. All students taking the course are automatically subscribed to this list. Previous postings can be accessed using the mailing list archive.