Planning to divide a big single collection to multiple collections to use $search [Lucene text analyser]

Hi, I am planning to break a big single collection ~10GB into multiple smaller collection correspond different user in application.

I recently implemented a search analyser of type autocomplete, but then I released that I don’t need for whole collection. Infact, using $search and normal userid index would return me the result in same time.

What I planned is to break one single collection into multiple collection, post that I can enable autocomplete text analyser to specific collection that I want for faster result.

Any suggestion will text analyser/search index in a single big collection is same in terms of performance as compared to multi collection approach with text analyser implemented in selected collection only ?

1 Like