Hello,
For my project I have an index with 1.000.000 documents.
Each people could add document to their favorites.
Now if somebody want to make a search in their favorites with facets,
there is 2 solutions:
- I create an index for favorites.
I get all favorites ids for the user.
And I make a query with "terms" for find all documents.
I made a test on my laptop for 10.000 favorites / 7sec. Maybe it's
faster on a server? Because 7 sec it's not possible.
Or second solution:
- I create a child index for favorites.
And I made a query with has_child or top_children for find all
documents.
But is it a good solution if the index become very large with the
time?
Somebody could advise me? thanks.
Another question: 90% percent of my data need to be indexed. So I
would like to know if it's possible to use only ES for stock all my
data, and is it SURE? thanks