Re: Default setting for enable_hashagg_disk

Peter Geoghegan <pg@bowt.ie>

From: Peter Geoghegan <pg@bowt.ie>
To: Alvaro Herrera <alvherre@2ndquadrant.com>
Cc: Jeff Davis <pgsql@j-davis.com>, David Rowley <dgrowleyml@gmail.com>, Stephen Frost <sfrost@snowman.net>, Tom Lane <tgl@sss.pgh.pa.us>, Andres Freund <andres@anarazel.de>, Bruce Momjian <bruce@momjian.us>, Robert Haas <robertmhaas@gmail.com>, Justin Pryzby <pryzby@telsasoft.com>, Melanie Plageman <melanieplageman@gmail.com>, Tomas Vondra <tomas.vondra@2ndquadrant.com>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2020-07-27T20:01:11Z
Lists: pgsql-hackers, pgsql-docs
On Mon, Jul 27, 2020 at 12:52 PM Alvaro Herrera
<alvherre@2ndquadrant.com> wrote:
> On 2020-Jul-27, Peter Geoghegan wrote:
> > The v4-0001-Remove-hashagg_avoid_disk_plan-GUC.patch changes are
> > surprisingly complicated. It would be nice if you could take a look at
> > that aspect (or confirm that it's included in your review).
>
> I think you mean "it replaces surprisingly complicated code with
> straightforward code".  Right?  Because in the previous code, there was
> a lot of effort going into deciding whether the path needed to be
> generated; the new code just generates the path always.

Yes, that's what I meant.

It's a bit tricky. For example, I have removed a redundant
"cheapest_total_path != NULL" test in create_partial_grouping_paths()
(two, actually). But these two tests were always redundant. I have to
wonder if I missed the point. Though it seems likely that that was
just an accident. Accretions of code over time made the code work like
that; nothing more.

-- 
Peter Geoghegan



Commits

  1. Add hash_mem_multiplier GUC.

  2. HashAgg: use better cardinality estimate for recursive spilling.

  3. Remove hashagg_avoid_disk_plan GUC.

  4. Doc fixup for hashagg_avoid_disk_plan GUC.

  5. Rework HashAgg GUCs.

  6. Disk-based Hash Aggregation.

  7. Implement partition-wise grouping/aggregation.

  8. Defer creation of partially-grouped relation until it's needed.