Database for a search engine

I am thinking of developing a search engine, but have no idea about the backend that i could use efficiently. Please suggest me a database in which i can store thousands 开发者_如何学Pythonof records and query them in a time efficient manner. I am developing search engine for my own interest so, please dont give me any critics thoughts as usually found in similar question.

Thankyou

Take a look at Cloudant, which is a clustered/scalable deployment of Apache's CouchDB. CouchDB is part of the NoSQL initiative (it uses the filesystem instead of having storage abstracted by a database server) and is told to scale quite well, especially with Cloudant, though I haven't tried it myself.

Obviously with any DODB (document-oriented database, as Apache puts it) you're going to have to shift your concerns from solely memory and table indexing to include storage capacity, though using the filesystem means you're generally able to use less memory (as less information is stored in memory). That's a broad generalization, but you get the idea.

For thousands of records, a standard rdbms will suffice. For 100's of millions and beyond, you might want to look into alternatives.

But since you are doing it for your own interest, do one with each so you understand the differences.

MySQL and Full-text indexes are a good solution to search inside thousands of records. When you scale up and get closer to 1 million rows, MySQL full text becomes slow and you will have to use another approach, like Sphinx.