Postgres has new release 9.6 - I have read the docs about the index-only scan. I'm aware of postgres tables' heap, and I know how partial indexes(or indexes in general) work. I just don't understand the benefit of index-only.
Questions
- What really changed? Didn't we used index only scan already?
- How and when should I use them to get the best benefits?
non related background:
I'm dealing with tables with 6B rows with a lot of INSERTs and SELECTs in production environments