Horton

Re: Default setting for enable_hashagg_disk

Tomas Vondra <tomas.vondra@2ndquadrant.com>

View thread View on postgr.es
From: Tomas Vondra <tomas.vondra@2ndquadrant.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Jeff Davis <pgsql@j-davis.com>, Peter Geoghegan <pg@bowt.ie>, Alvaro Herrera <alvherre@2ndquadrant.com>, David Rowley <dgrowleyml@gmail.com>, Stephen Frost <sfrost@snowman.net>, 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>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2020-07-20T17:25:39Z
Lists: pgsql-hackers, pgsql-docs 
On Mon, Jul 20, 2020 at 09:17:21AM -0400, Tom Lane wrote:
>Tomas Vondra <tomas.vondra@2ndquadrant.com> writes:
>> There's a minor problem here, though - these stats were collected before
>> we fixed the tlist issue, so hashagg was spilling about 10x the amount
>> of data compared to sort+groupagg. So maybe that's the first thing we
>> should do, before contemplating changes to the costing - collecting
>> fresh data. I can do that, if needed.
>
>+1.  I'm not sure if we still need to do anything, but we definitely
>can't tell on the basis of data that doesn't reliably reflect what
>the code does now.
>

OK, will do. The hardware is busy doing something else at the moment,
but I'll do the tests and report results in a couple days.


regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

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.