Sneak peek photo [1] (from [2]). Just imagine its literally 500-1000x more convoluted per B-tree leaf node. That's every Postgres table unless you CLUSTER periodically.
[1]: https://josipmisko.com/img/clustered-vs-nonclustered-index.w...
[2]: https://josipmisko.com/posts/clustered-vs-non-clustered-inde...
Mind boggling how many people aren't aware of primary indexes in MySQL that is not supported at all in Postgres. For certain data layouts, Postgres pays either 2x storage (covering index containing every single column), >50x worse performance by effectively N+1 bombing the disk for range queries, or blocking your table periodically (CLUSTER).
In Postgres the messiness loading primary data after reaching the B-tree leaf nodes pollutes caches and takes longer. This is because you need to load one 8kb page for every row you want, instead of one 8kb with 20-30 rows packed together.
Example: Dropbox file history table. They initially used autoinc id for primary key in MySQL. This causes everybodys file changes to be mixed together in chronological order on disk in a B-Tree. The first optimization they made was to change the primary key to (ns_id, latest, id) so that each users (ns_id) latest versions would be grouped together on disk.
Dropbox scaling talk: https://youtu.be/PE4gwstWhmc?t=2770
If a dropbox user has 1000 files and you can fit 20 file-version rows on each 8kb disk page (400bytes/row), the difference in performance for querying across those 1000 files is 20 + logN disk reads (MySQL) vs 1000 + logN disk reads (Postgres). AKA 400KiB data loaded (MySQL) vs 8.42MiB loaded (Postgres). AKA >50x improvement in query time and disk page cache utilization.
In Postgres you get two bad options for doing this: 1) Put every row of the table in the index making it a covering index, and paying to store all data twice (index and PG heap). No way to disable the heap primary storage. 2) Take your DB offline every day and CLUSTER the table.
Realistically, PG users pay that 50x cost without thinking about it. Any time you query a list of items in PG even using an index, you're N+1 querying against your disk and polluting your cache.
This is why MySQL is faster than Postgres most of the time. Hopefully more people become aware of disk data layout and how it affects query performance.
There is a hack for Postgres where you store data in an array within the row. This puts the data contiguously on disk. It works pretty well, sometimes, but it’s hacky. This strategy is part of the Timescale origin story.
Open to db perf consulting. email is in my profile.