The performance of query records type

Question

I have a question about the performance of data processing. 
 I have a query a list from record type A, and now I want to filter some data from the data I queried. I have two ways to do that, one is use query record type and add a filter to get the data I want, another is I use logic function to get the data from the list I queried. My question is, is there a big performance difference between these two solutions? If yes which way is preferred. If no how about CDT. Because when I code in Java, our strategy is better to query the database only once. But I'm not sure is this suit for Appian. But I think records type has cache like Redis, so even query the record type too much times, it will not cause performance problems. But I'm not sure if my thought is right?

Shubham Aware · Answer

I would recommend always using a!queryRecordType() with filters to fetch only needed data, as it pushes filtering to the database layer for optimized queries. Filtering large lists in memory is slower and not preferred. This matches your Java strategy: query the database efficiently once with proper filters. Use in-memory filtering only for small datasets.

Harsha Sharma · Answer

Querying with filters applied is better so that the query returns just the data needed for the usecase. It reduces the 'extra' data being queried, which you might not need eventually, after further filtering anyway. Also, functions can slow down filtering while working with large dataset. So I suggest to use filters when querying large datasets. 
 If the query returns small dataset, then depending on reusability opportunities or various manipulations needed, functions like index(), property() etc are more suitable instead of multiple queries! Due to small data size performance and memory impacts risks are reduced.

Stefan Helzle · Answer

Filtering in the DB is almost always the better option.