Re: Planning time is time-consuming

From: Imre Samu <pella(dot)samu(at)gmail(dot)com>
To: Mikhail Balayan <mv(dot)balayan(at)gmail(dot)com>
Cc: pgsql-performance(at)postgresql(dot)org
Subject: Re: Planning time is time-consuming
Date: 2023-09-11 15:17:33
Message-ID: CAJnEWwn-z1Ef96+n5LvNQoTh53FjxYGh3GzDGegvbEjJd5y4jw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Hi Mikhail.

Postgresql version: 15.3 (Debian 15.3-1.pgdg110+1) on x86_64-pc-linux-gnu,
> compiled by gcc (Debian 10.2.1-6) 10.2.1 20210110, 64-bit
> And just in case it matters, this is an experimental setup, so Postgresql
> running in Docker.
>

Are you using the official Docker Postgres image, specifically
`postgres:15.3-bullseye`? ( https://hub.docker.com/_/postgres )
- If so, consider upgrading to version 15.4. It has some planner fixes not
directly related to your issue. Check details here:
PostgreSQL 15.4 Release Notes
https://www.postgresql.org/docs/release/15.4/
- For all technical text type id columns *apply the `Collate "C"` *. (
like `assign_time_str` and `cti_entity_uuid` )
Alternatively, use the "uuid" column type everywhere, as Tom Lane
suggests.

- Could you provide details on your current tuning settings? I'm interested
in `work_mem`, `shared_buffers`, `effective_cache_size`, and others.
- Please test with different `work_mem` values.

If it's not too much trouble, can you also test with: ( These version uses
a different locale and LLVM (JIT). )
- postgres:15.4-bookworm
- postgres:15.4-alpine3.18

Regards,
Imre

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message David Rowley 2023-09-12 03:06:12 Re: Planning time is time-consuming
Previous Message Tom Lane 2023-09-11 14:27:13 Re: Planning time is time-consuming