DBSight is a high performance and flexible search engine for any databases. In addition to normal full text search, it also provides multiple varieties of facet search, scaffolding, sharding, distributed search. It also has integrated SQL crawler and API for data beyond databases.

Free DBSight Lite enables users with basic SQL knowledge to create scalable Lucene search on any database, with arrays of features like Time-based ranking, facet-search, integrated spell checking, suggest-as-you-type, and several pre-built scaffoldings to bootstrap from.

You can create a database search service, return results as HTML/XML/JSON/JSONP, so you can easily use it with Ruby, PHP, or any existing web applications. Or you can use the java API to search, and return results in an open-sourced format, defined by the efficient Google Protocol Buffer.  It's easily customizable. You can add your own analyzer, similarity for your special needs.
It's also highly scalable. You can create sharded indexes for your sharded databases, and have a seamless global search.

The best thing is, it's free and super easy. You can create a production-level search in 3 minutes.

By making database full-text search super easy to create or change, you can quickly add a new full-text search or adjust current index structure according to your specific requirements.  DBSight makes database searching a detachable separate concern, greatly improves the flexibility and scalability. DBSight also does the "dirty" job to synchronize the database, incremental indexing, facet-search(categorized result counting), easy to change synonyms and stop word, reserved word list, field-level and document-level ranking, multi-threaded searching, time-based ranking, transacdtional index data integrity, server clustering, etc.
DBSight powers production-level sites. For example, one product-search site powered by DBSight recently got funded 2.6 million Euros. You can do it too.