From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Erik Nordström <erik(at)timescale(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Missing Group Key in grouped aggregate |
Date: | 2024-02-20 15:53:58 |
Message-ID: | 185735.1708444438@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
=?UTF-8?Q?Erik_Nordstr=C3=B6m?= <erik(at)timescale(dot)com> writes:
> I noticed that, beginning with PG16, grouped aggregates are missing the
> "Group Key" in the EXPLAIN output.
> It seems the Agg node has numCols (number of grouping cols) set to zero in
> queries like
> SELECT foo, count(*) FROM bar WHERE foo=1 GROUP BY foo;
> In PG15, the "Group Key" is shown and the Agg node has numCols set as
> expected.
Looks sane to me: the planner now notices that there can only
be one group so it doesn't tell the GroupAgg node to worry about
making groups. If it were missing in a case where there could be
multiple output groups, yes that'd be a bug.
If you want to run it to ground you could bisect to see where the
behavior changed, but you'd probably just find it was intentional.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Bertrand Drouvot | 2024-02-20 15:55:08 | Re: Injection points: some tools to wait and wake |
Previous Message | Tom Lane | 2024-02-20 15:45:51 | Re: Change the bool member of the Query structure to bits |