I actually have 200 GB of data indexed on MongoDB, spread over 2 columns, each containing one string.
The user search for 'cowboy', if this string is found anywhere, it returns the content.
i.e: Description / Name
'cowboyis-xz' / 'foo' 'lorem lipsum..cowboy.' / 'bar' -> FOUND: 'cowboyis-xz' / 'toto' -> FOUND: 'lorem lipsum..cowboy.' / 'bar'
I'm looking for the most efficient way to make a full text search with a DB respecting those criteria:
- Fastest research with any pattern (so with a minimum of complexity)
- Through dozens of billions of row (~ 500 GB to 2 TB of data)
- With fixed data structure (2 columns, maybe 3 or 4 in the future)
- With fast data imports
What would be the fastest way to make full text searches ?
I'm comparing those technologies :
- ElasticSearch / Sphinx
- "Text" index on MongoD
FULL TEXTfunctionality on MySQL
If anyone has already tried / benchmarked those DB & search engines for that kind of needs, I would really appreciate your help.