Database Reference
In-Depth Information
izing it to disk like MapReduce does. Being able to try a pipeline on different engines
without rewriting the program is a powerful property, since it allows you to treat what the
program does separately from matters of runtime efficiency (which generally improve
over time as the engines are tuned).
This chapter is an introduction to writing data processing programs in Crunch. You can
find more information in the
Crunch User Guide
.
