Day 2: Mapreduce and Server Clusters
Today we’ll dive into the mapreduce framework to perform more powerful queries than the standard key-value paradigm can normally provide. We’ll then expand on this power by including link walking with mapreduce. Finally, we will investigate the server architecture of Riak and how it uses a novel server layout to provide flexibility in consistency or availability, even in the face of network partitions.
We’ll need a bit more data in this section. To achieve that, we’ll switch to an example using a different kind of hotel, one for people and not pets. A quick populator script in Ruby will create data for a gigantic 10,000-room hotel.
If you are not familiar with Ruby, it is a popular general-purpose ...