Re: strange slow query - lost lot of time somewhere - Mailing list pgsql-hackers

From Justin Pryzby
Subject Re: strange slow query - lost lot of time somewhere
Date
Msg-id 20220510022248.GU28830@telsasoft.com
Whole thread Raw
In response to Re: strange slow query - lost lot of time somewhere  (David Rowley <dgrowleyml@gmail.com>)
List pgsql-hackers
On Tue, May 03, 2022 at 02:13:18PM +1200, David Rowley wrote:
> I'm wishing I put the initial hash table size and the final hash table
> size in EXPLAIN + EXPLAIN ANALYZE now. Perhaps it's not too late for
> v15 to do that so that it might help us figure things out in the
> future.

On Fri, May 06, 2022 at 09:27:57PM +1200, David Rowley wrote:
> I'm very tempted to change the EXPLAIN output in at least master to
> display the initial and final (maximum) hash table sizes. Wondering if
> anyone would object to that?

No objection to add it to v15.

I'll point out that "Cache Mode" was added to EXPLAIN between 11.1 and 11.2
without controversy, so this could conceivably be backpatched to v14, too.

commit 6c32c0977783fae217b5eaa1d22d26c96e5b0085
Author: David Rowley <drowley@postgresql.org>
Date:   Wed Nov 24 10:07:38 2021 +1300

    Allow Memoize to operate in binary comparison mode



pgsql-hackers by date:

Previous
From: "Jonathan S. Katz"
Date:
Subject: Re: 2022-05-12 release announcement draft
Next
From: Richard Guo
Date:
Subject: Re: psql now shows zero elapsed time after an error