Impact of schema fields limit on performance

Elastic Search
1

Hi there,

I'm currently using App Search with search-ui react library.

I was wondering about the search performance impact of increasing the schema fields limit (for ex. 70 vs 100 fields -- only 10 fields are used for relevance tuning in both cases). Will there be a noticeable difference in search response times or will it be negligible if I maintain the same fields for relevance tuning?

Thanks!

2

Thanks for your question!

Do you mean that you're only searching over 10 fields, and returning an additional 30, or that you search over 70-100 fields but only 10 have a specific relevance tuning?

If it's the first case, the performance impact should be negligible.

If it's the latter, it will really depend. Here are a few things that can impact performance:

  • The shape of the data (e.g. the size of the fields and values)
  • What sort of search you're performing (e.g. if you're faceting over some high-cardinality fields)
  • The type of response you want (e.g. if you want highlighting enabled for all those fields) – relevance settings should help here.

But ultimately the best way to check would be to just try it out!

Hope that helps – please let us know if you have any follow up questions.

Nick

1 Like
3

Thanks for your response Nick!

I was referring to the "app_search.engine.total_fields.limit" setting, let me see if I can break it down:

Scenario 1:
100 fields (stored but not searched)
10 fields searched and used for relevance tuning
5 fields returned

Scenario 2:
800 fields (stored but not searched)
10 fields searched and used for relevance tuning
5 fields returned

I think the performance be about the same since I'm searching over the same amount of fields?

I'll try it out! Thanks for the tips, I'll keep them in mind :+1:

4

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