3n+1 NoSQL Database Benchmark

During my research I stumbled over this interesting analysis and benchmark.It's an attempt to create a benchmark for databases that are generally described as NoSQL, key-value, schema-free or schema-less databases. The goals are to create a benchmark that: * is simple to understand and code - ideally, a page of code or less in most programming languages; * performs a non-trivial amount of representative activity - reads and updates, access contention, recovery, etc.; and * generates results that can be easily replicated and verified by others.