Message-ID: <4f462695-5716-408a-a08a-8ba5e665bb9a@gmail.com>
Date: Tue, 13 Jan 2026 10:29:45 +0100
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: small database huge planning time
To: Alexander Kulikov <a-kulikov@hotmail.com>,
 pgsql-performance@postgresql.org
References: 
 <AS4PR02MB874303264FDEF9B160C06DF9EA8EA@AS4PR02MB8743.eurprd02.prod.outlook.com>
Content-Language: en-US
From: Andrei Lepikhov <lepihov@gmail.com>
In-Reply-To: 
 <AS4PR02MB874303264FDEF9B160C06DF9EA8EA@AS4PR02MB8743.eurprd02.prod.outlook.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Archived-At: 
 <https://www.postgresql.org/message-id/4f462695-5716-408a-a08a-8ba5e665bb9a%40gmail.com>
Precedence: bulk

On 13/1/26 10:16, Alexander Kulikov wrote:
> Hello!
> 
> 
> I have got huge planning time for a query in quite small database in PortgreSQL 17
>   Planning Time: 452.796 ms
>   Execution Time: 0.350 ms
> 
> Tried several version from 17.3 to 17.7 (cpu 2.2GHz) - it almost does not matter. If I run query many times in row planning time may reduce down to 430ms but never less.
> 
> Tried in PortgreSQL 11 (in a little bit different hardware with cpu 2.60GHz) - planning time almost ten times less.
> 
> Changing parameters: from_collapse_limit, join_collapse_limit, geqo, jit, work_mem and many 
> others does not help at all. I attach 1. additional setting in the 
> postgresql.status.conf. 2. querry itself in query.sql. 3. zql plan in 
> query.sqlplan 4. additioanal information about os, tables etc. would you 

May you attach the same EXPLAIN for the 'good' execution? Is there any 
chance to get the schema definition as a SQL script (ideally, 
without/replaced mvarchar fields)? Is it reproducible on a database with 
empty tables?

-- 
regards, Andrei Lepikhov,
pgEdge