Cost of searchable fields

I'm very new to Elastic. I am thinking of a document structure of a reasonable size, with quite a few fields, which I don't necessarily need to be able to search on. What I could not find in the documentation is the exact impact of having too many searchable fields. For eg. with a certain number of documents, what is the storage/performance impact of having 10 searchable fields vs say, 100? Any insight and best practices in this regard will be great.

Hi @abhishekc92 Welcome to the community.

Great question and of course with any system there are some tradeoffs, but as you say you are just starting your journey. My advice with Elasticsearch is to not try to over optimize when you are getting started. The difference between 10 and 100 indexed fields could be similar or quite different depending on a lot of factors that are too numerous to list in a single post.

Elasticsearch is designed to hold large / massive quantities of documents / data Millions, 100s of Millions, Billions of documents (small and large), so unless you are starting with some extreme volumes day I would not try to over optimize. Just index the documents. Elastic is very efficient with its data structures, and of course if you have small data set, it is good with those too.

If you are coming from an RDBMS background the amount of data elasticsearch can easily handle can seem... enormous.

What I would do if I am starting is I would focus on good understanding of Mappings (schema), Templates, Indexing and Searches.

I would start with some of our free training video and Free Training

All that said you can experiment, there is a whole section on tuning for disk usage here

Come back with more questions as you start to experiment more...


This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.