Learning objectives:

In this lecture, you will learn:

• Lec3a: the design and architecture of GFS, data replication mechanisms, how GFS serves reads and writes.
• Lec3b: the design, the programming model, the workflow of typical MapReduce applications (Word Count), and the fault tolerance model of MapReduce.
• Putting it all together: how the MapReduce computing framework interacts with the underlying GFS.

Lecture slides

• Lec3a: GFS: slides pdf, slides+notes
• Lec3b: MapReduce: slides pdf, slides+notes

Readings

• Lec3a: The GFS paper
• Lec3b: The MapReduce paper

Paper review forms

• GFS: form
• MapReduce: form