Course: B534 Distributed systems Graduate/Undergraduate Class
Abstract
A class of 60 students (PhD, Masters, Undergraduates) covering core Computer Science distributed system curricula.
Intellectual Merit
Class Information
Time: MW 2:30PM - 3:45PM
Place: Informatics East (I2) 150
Bloomington, IN 47405
Office Hours
Instructor: Prof. Judy Qiu
AIs: Lkhyun Park, Pairoj Rattadilok
Friday 3:00pm to 4:00pm
Lindley Hall Room 201D
Prerequisites
CSCI-P 436 or P536 is required (or permission from the instructor). General programming experience with Windows or Linux using Java, C#, or C++, scripts is preferred. Parallel and cluster computing background is a plus although not required.
Objectives
The Internet has greatly expanded the scope and importance of distributed systems to include Web 2.0 sites, Information retrieval (search), Utility (cloud) computing, P2P systems and the Internet of things. Further science is facing an unprecedented data deluge and the emergence of data oriented analysis as a fourth paradigm of scientific methodology after theory, experiment and simulation. This class will use these modern systems to introduce core technologies including communication, concurrency/parallelism, security, fault tolerance and programming models. In particular the course will cover programming models and tools of cloud computing to support data intensive science applications. Students will get to know the latest research topics of cloud platforms and have the opportunity to understand some commercial cloud systems through projects using FutureGrid resources.
Scope and topcis
The content of B534 will cover the design principles, systems architecture, and innovative applications of parallel, distributed, and cloud computing systems. These include massively parallel processors (MPP), supercomputing clusters, service-orient architecture (SOA), computational grids, P2P (peer-to-peer) networks, virtualized datacenters, cloud platforms, Internet of Things (IOT), and Cyber-Physical Systems (CPS). We will cover MapReduce (originated from functional language) and associated languages like Sawzall and DryadLINQ but it will focus more on the principles and practice of building distributed systems than on languages.
Broader Impact
The curricula and tutorials can be re-used in other cloud computing/distributed system educational activities
Use of FutureGrid
Support 70 students in distributed system class
Scale Of Use
Each student will need modest resources