Question

I am building a tagging system (think post tagging in a blog) that will use two tables in mysql.

First table will have:

• tag_id (int)
• tag (varchar)

Second table will have:

• tag_id
• post_id (to link them)

When adding a tag, the first thing I want to do is check if the tag already exists in the first table.

How to do that in the most efficient way? Should I just do

SELECT tag_id from tags where tag = 'atag'

If so, what is the best way to index the tag field?

Is it more efficient if I create a third field with a hash of the tag and index and search that?

I expect the number of tags to grow into the hundreds of thousands.

Answer 1

Since

• tags are relativly short
• you want tags to be unique

I’d say a normal unique index on tags.tag is the way to go.

Additionally, since a few tags will account for a big part of the tag cloud you might want to consider LRU-caching them in RAM.