Horton

Re: BitmapHeapScan streaming read user and prelim refactoring

Jakub Wartak <jakub.wartak@enterprisedb.com>

View thread View on postgr.es
From: Jakub Wartak <jakub.wartak@enterprisedb.com>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Tomas Vondra <tomas@vondra.me>, Melanie Plageman <melanieplageman@gmail.com>, Nazir Bilal Yavuz <byavuz81@gmail.com>, Dilip Kumar <dilipbalaut@gmail.com>, Heikki Linnakangas <hlinnaka@iki.fi>, Thomas Munro <thomas.munro@gmail.com>, Andres Freund <andres@anarazel.de>, Pg Hackers <pgsql-hackers@postgresql.org>
Date: 2025-02-14T09:04:41Z
Lists: pgsql-hackers

Attachments

On Wed, Feb 12, 2025 at 9:57 PM Robert Haas <robertmhaas@gmail.com> wrote:
>
> On Wed, Feb 12, 2025 at 3:07 PM Tomas Vondra <tomas@vondra.me> wrote:
> > AFAICS the "1" value is simply one of the many "defensive" defaults in
> > our sample config. It's much more likely to help than cause harm, even
> > on smaller/older systems, but for many systems a higher value would be
> > more appropriate. There's usually a huge initial benefit (say, going to
> > 16 or 32), and then the benefits diminish fairly quickly.
>
> I'm happy to see us change the value to something that is likely to be
> good for most people. I think it's OK if people on very tiny systems
> need to change a few defaults for optimum performance. We should keep
> in mind that people do sometimes run PostgreSQL on fairly small VMs
> and not go crazy with it, but there's no reason to pretend that the
> typical database runs on a Raspberry Pi.

Is there any reason we couldn't have new pg_test_iorates (similiar to
other pg_test_* proggies), that would literally do this and calibrate
best e_io_c during initdb and put the result into postgresql.auto.conf
(pg_test_iorates --adjust-auto-conf) , that way we would avoid user
questions on how to come with optimal value?

root@jw-test3:/nvme# ./pg_test_iorates
File size: 16.00 GB, Block size: 8192 bytes
buffered sequential read: 1573.33 MB/s
direct sequential read: 289.46 MB/s
random read performance with different prefetch distances:
using prefetch distance 1: 173.58 MB/s
using prefetch distance 2: 243.05 MB/s
using prefetch distance 4: 376.78 MB/s
using prefetch distance 8: 590.66 MB/s
using prefetch distance 16: 824.49 MB/s
using prefetch distance 32: 861.45 MB/s
using prefetch distance 64: 830.36 MB/s

Attached, compile naive way via: gcc pg_test_iorates.c -o
pg_test_iorates -I /git/postgres/build/src/include -I
/git/postgres/src/include -L /git/postgres/build/src/common
-L/git/postgres/build/src/port -lpgcommon -lpgport -lm

-J.

Commits

  1. Fix bitmapheapscan incorrect recheck of NULL tuples

  2. Increase default maintenance_io_concurrency to 16

  3. Separate TBM[Shared|Private]Iterator and TBMIterateResult

  4. Improve read_stream.c advice for dense streams.

  5. Increase default effective_io_concurrency to 16

  6. Delay extraction of TIDBitmap per page offsets

  7. Add lossy indicator to TBMIterateResult

  8. Move BitmapTableScan per-scan setup into a helper

  9. Add and use BitmapHeapScanDescData struct

  10. Fix bitmap table scan crash on iterator release

  11. Bitmap Table Scans use unified TBMIterator

  12. Add common interface for TBMIterators

  13. Make table_scan_bitmap_next_block() async-friendly

  14. Move EXPLAIN counter increment to heapam_scan_bitmap_next_block

  15. Refactor tidstore.c iterator buffering.

  16. BitmapHeapScan: Remove incorrect assert and reset field

  17. Change BitmapAdjustPrefetchIterator to accept BlockNumber

  18. BitmapHeapScan: Use correct recheck flag for skip_fetch

  19. BitmapHeapScan: Push skip_fetch optimization into table AM

  20. BitmapHeapScan: postpone setting can_skip_fetch

  21. BitmapHeapScan: begin scan after bitmap creation

  22. Fix EXPLAIN Bitmap heap scan to count pages with no visible tuples

  23. Remove redundant snapshot copying from parallel leader to workers

  24. Remove some obsolete smgrcloseall() calls.

  25. Remove the "snapshot too old" feature.

  26. Compute XID horizon for page level index vacuum on primary.