I've been following elasticsearch for a while now - it's a fascinating project. It's similar to Solr, but it's designed from the ground up for seamless sharded horizontal scaling (Solr is mainly designed for master/slave replication), speakers JSON/REST by default and keeps on getting more and more interesting features - geospatial support is baked in for example.
Percolator has really captured my imagination. Instead of storing documents and running searches against them, you can store searches and feed in a document and ask "which search queries would this match". It turns search upside down. I'm not a hundred percent sure what this is useful for (two examples would be a saved-search alert feature of some sort or something for firing off events based on an incoming stream of data, such as the Twitter firehose) but it's a fascinating ability.
Another use case would be to setup business rules or even get a workflow going. If i add this document it matches the queries a,b,c so i can do steps x,y and/or z afterwards.
Percolators is a truely perculiar feauture indeed :)
Percolator has really captured my imagination. Instead of storing documents and running searches against them, you can store searches and feed in a document and ask "which search queries would this match". It turns search upside down. I'm not a hundred percent sure what this is useful for (two examples would be a saved-search alert feature of some sort or something for firing off events based on an incoming stream of data, such as the Twitter firehose) but it's a fascinating ability.