Big Data Computing

From Hadoop to key-value stores

The course moves from big data fundamentals into the systems that make it practical. It starts with enabling technologies and the Hadoop stack, then looks at HDFS and MapReduce in its 1.0 and 2.0 forms, before introducing Spark and its built-in libraries.

The second half turns to storage and distributed design: key-value store structure, data placement strategies, the CAP theorem and consistency solutions. Those topics suit readers who want to understand trade-offs, not only tools.

Who it is for

Computer science students and engineers who need to understand how big data systems are built will find the lecture order logical. It is also suited to anyone who wants a grounding in distributed storage before choosing a platform.

Points to consider

• Work through the MapReduce lectures carefully, since the Spark material builds on them.
• Use the lecture list to plan your study time across the 34 lectures.