On Thu, Apr 11, 2019 at 10:28 AM rihad <rihad@mail.ru> wrote:
Yup, it's just that n_dead_tuples grows by several hundred thousand (the table sees much much more updates than inserts) and disk usage grows constantly between several hour long vacuum runs. Running vacuum full isn't an option.
The disk usage doesn't reach a steady state after one or two autovacs? Or it does, but you are just unhappy about the ratio between the steady state size and the theoretical fully packed size?