Motivations

OpenStreetMap contributors are working hard to collect, store and maintain gigabytes of valuable geographical data (look at there if you're not yet convinced!). Among the tools created to serve these common goals, Osmosis is a powerful ETL-like Java application that eases importing / exporting / manipulate these data in a PostgreSQL relational database.

However, problems may appear when you want to query such large datasets under high load: RDBMS are known to not scale very well in such scenarios. In addition, you are probably not interested in the richness of the OpenStreetMap data model or your needs are just simpler than what it offers.

Fortunately, since the rise of NoSQL databases, we now have other storage options. elasticsearch is one of these. This promising search engine leverages the Java Lucene API to offer fast data indexing and querying over a distributed cluster. In addition - and that's why we are here - it offers some greatful geo search features.

I started this project with the feeling of a natural combination between geographical data, a specialized ETL tool and a geo-enabled search engine.