Yeah, we call it Incremental MapReduce and we _LOVE_ it. I swear, it's one of the biggest "ah ha!" moments whenever I do training for Cloudant's customers.
The basic idea is that the intermediate Map function results are persisted, while most other system (read: hadoop) throws them out.
The basic idea is that the intermediate Map function results are persisted, while most other system (read: hadoop) throws them out.
If you're interested in some of the MapReduce history, where it's going, incremental indexing, etc., then you should read this: http://gigaom.com/cloud/why-the-days-are-numbered-for-hadoop...
Cheers.