Re: Which partition scheme makes sense for my time based IoT-datagrams?

On Wed, 2021-02-10 at 18:59 +0100, Thorsten Schöning wrote:
> > They will scan all partitions. You have to be prepared that most queries
> > will become at least slightly slower with partitioning. That is
> > expected.
>
> Does "most queries" address thos efor IDs scanning all partitions or
> those time-based as well? In the end, I'm trying to improve query
> performance by reducing the size of indexes, number of rows etc. per
> table using partitions. :-)

The size of the index will barely affect its performance.

But having to figure out which partitions to scan takes planning time.

Very few queries become notably faster through partitioning.
Mostly those are aggregates and joins that match the partitioning scheme,
and sequential scans that include the partitioning key.

Yours,
Laurenz Albe
--
Cybertec | https://www.cybertec-postgresql.com