From: | Christophe Pettus <xof(at)thebuild(dot)com> |
---|---|
To: | pgsql-performance(at)lists(dot)postgresql(dot)org |
Cc: | Andrew Dunstan <andrew(at)dunslane(dot)net>, Jan Wieck <jan(at)wi3ck(dot)info>, "Dean Gibson (DB Administrator)" <postgresql(at)mailpen(dot)com>, Bruce Momjian <bruce(at)momjian(dot)us> |
Subject: | Re: AWS forcing PG upgrade from v9.6 a disaster |
Date: | 2021-05-29 02:43:23 |
Message-ID: | E007EEB7-097D-4D0E-9C4D-C8BB28810E7F@thebuild.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers pgsql-performance |
> On May 28, 2021, at 14:30, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> I think it uses pg_upgrade.
It does. It does not, however, do the vacuum analyze step afterwards. A VACUUM (FULL, ANALYZE) should take care of that, and I believe the OP said he had done that after the pg_upgrade.
The most common reason for this kind of inexplicable stuff after an RDS upgrade is, as others have said, parameter changes, since you get a new default parameter group after the upgrade.
That being said, this does look like something happened to the planner to cause it to pick a worse plan in v13. The deeply nested views make it kind of hard to pin down, but the core issue appears to be in the "good" plan, it evaluates the _Club.club_count > 5 relatively early, which greatly limits the number of rows that it handles elsewhere in the query. Why the plan change, I can't say.
It might be worth creating a materialized CTE that grabs the "club_count > 5" set and uses that, instead of having it at the top level predicates.
From | Date | Subject | |
---|---|---|---|
Next Message | Masahiko Sawada | 2021-05-29 02:56:39 | Re: Skipping logical replication transactions on subscriber side |
Previous Message | Andrew Dunstan | 2021-05-29 02:41:08 | Re: AWS forcing PG upgrade from v9.6 a disaster |
From | Date | Subject | |
---|---|---|---|
Next Message | Dean Gibson (DB Administrator) | 2021-05-29 04:08:28 | Re: AWS forcing PG upgrade from v9.6 a disaster |
Previous Message | Andrew Dunstan | 2021-05-29 02:41:08 | Re: AWS forcing PG upgrade from v9.6 a disaster |