Horton

Re: Trouble with hashagg spill I/O pattern and costing

Jeff Davis <pgsql@j-davis.com>

View thread View on postgr.es
From: Jeff Davis <pgsql@j-davis.com>
To: Tomas Vondra <tomas.vondra@2ndquadrant.com>
Cc: Thomas Munro <thomas.munro@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>, Melanie Plageman <melanieplageman@gmail.com>, HeikkiLinnakangas <hlinnaka@iki.fi>
Date: 2020-05-29T00:48:11Z
Lists: pgsql-hackers

Attachments

On Tue, 2020-05-26 at 17:40 -0700, Jeff Davis wrote:
> On Tue, 2020-05-26 at 21:15 +0200, Tomas Vondra wrote:
> > Yeah. I agree prefetching is definitely out of v13 scope. It might
> > be
> > interesting to try how useful would it be, if you're willing to
> > spend
> > some time on a prototype.
> 
> I think a POC would be pretty quick; I'll see if I can hack something
> together.

Attached (intended for v14).

I changed the list from a simple array to a circular buffer so that we
can keep enough preallocated block numbers in it to do prefetching.

On SSD I didn't see any improvement, but perhaps it will do better on
magnetic storage.

Regards,
	Jeff Davis

Commits

  1. Use CP_SMALL_TLIST for hash aggregate

  2. Avoid fragmentation of logical tapes when writing concurrently.