On Wed, Jun 03, 2020 at 09:45:48AM +0200, Oleksandr Shulgin wrote: > I see. So it works with low cardinality in the partitioned column. With > high cardinality an index scan on an unpartitioned table would be > preferable I guess. > > The documentation page I've linked only contains examples around > partitioning BY RANGE. I believe it'd be helpful to extend it with some > meaningful examples for LIST and HASH partitioning. I agree. I think it would also be useful to mention the "benefits" which aren't likely to apply to hash partitioning. Would you want to propose an example to include ? Eventually it needs to be submitted as a patch to -hackers. -- Justin
