iulian dragos writes:
> Is there any way to validate the bloated index hypothesis before I invest
> too much in this direction?
A plain old VACUUM ought to take care of most of the problem.
regards, tom lane
po 20. 12. 2021 v 17:50 odesÃlatel iulian dragos <
iulian.dra...@databricks.com> napsal:
> Thanks for the suggestion!
>
> Reading through the docs it looks like this might involve some downtime
> (locking writes to the tables whos indexes are being rebuilt) and this is a
> running system. I may ne
Thanks for the suggestion!
Reading through the docs it looks like this might involve some downtime
(locking writes to the tables whos indexes are being rebuilt) and this is a
running system. I may need some time to setup a database copy and try to
reproduce the issue (dev deployment doesn't exhibi
Hi
po 20. 12. 2021 v 13:31 odesÃlatel iulian dragos <
iulian.dra...@databricks.com> napsal:
> Hi,
>
> I was analyzing the query performance in a certain code path and noticed
> that practically all of the query time is spent planning (11s planning,
> 200ms execution time). Here is the output of E
