Highly scalable db schema for recurring events

Which is the best way to design a db schema in which store hundreds of thousand events, with recurrencies, and that will support milions of query asking which events will occur in a range of dates?

I mean, designing the tables to describe the model should be not difficult, but doing that in such a way that the huge amount of da开发者_Go百科ta can be searched quickly... well, I have no idea of how to do that.

Why not build your datamodel as you normally would (3NF-5NF) and add another layer that is better suited to handle searching like Lucene or Solr?