Horton

Re: [PoC] Improve dead tuple storage for lazy vacuum

Masahiko Sawada <sawada.mshk@gmail.com>

View thread View on postgr.es
From: Masahiko Sawada <sawada.mshk@gmail.com>
To: John Naylor <john.naylor@enterprisedb.com>
Cc: Nathan Bossart <nathandbossart@gmail.com>, Andres Freund <andres@anarazel.de>, Matthias van de Meent <boekewurm+postgres@gmail.com>, Yura Sokolov <y.sokolov@postgrespro.ru>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2023-02-22T08:29:23Z
Lists: pgsql-hackers 
On Wed, Feb 22, 2023 at 4:35 PM John Naylor
<john.naylor@enterprisedb.com> wrote:
>
>
> On Wed, Feb 22, 2023 at 1:16 PM Masahiko Sawada <sawada.mshk@gmail.com> wrote:
> >
> > On Mon, Feb 20, 2023 at 2:56 PM Masahiko Sawada <sawada.mshk@gmail.com> wrote:
> > >
> > > Yeah, I did a similar thing in an earlier version of tidstore patch.
>
> Okay, if you had checks against the old array lookup in development, that gives us better confidence.
>
> > > Since we're trying to introduce two new components: radix tree and
> > > tidstore, I sometimes find it hard to investigate failures happening
> > > during lazy (parallel) vacuum due to a bug either in tidstore or radix
> > > tree. If there is a bug in lazy vacuum, we cannot even do initdb. So
> > > it might be a good idea to do such checks in USE_ASSERT_CHECKING (or
> > > with another macro say DEBUG_TIDSTORE) builds. For example, TidStore
> > > stores tids to both the radix tree and array, and checks if the
> > > results match when lookup or iteration. It will use more memory but it
> > > would not be a big problem in USE_ASSERT_CHECKING builds. It would
> > > also be great if we can enable such checks on some bf animals.
> >
> > I've tried this idea. Enabling this check on all debug builds (i.e.,
> > with USE_ASSERT_CHECKING macro) seems not a good idea so I use a
> > special macro for that, TIDSTORE_DEBUG. I think we can define this
> > macro on some bf animals (or possibly a new bf animal).
>
>  I don't think any vacuum calls in regression tests would stress any of this code very much, so it's not worth carrying the old way forward. I was thinking of only doing this as a short-time sanity check for testing a real-world workload.

I guess that It would also be helpful at least until the GA release.
People will be able to test them easily on their workloads or their
custom test scenarios.

Regards,

-- 
Masahiko Sawada
Amazon Web Services: https://aws.amazon.com

Commits

  1. radixtree: Fix SIGSEGV at update of embeddable value to non-embeddable.

  2. Get rid of anonymous struct

  3. Teach radix tree to embed values at runtime

  4. Teach TID store to skip bitmap for small numbers of offsets

  5. Use bump context for TID bitmaps stored by vacuum

  6. Fix alignment of stack variable

  7. Use TidStore for dead tuple TIDs storage during lazy vacuum.

  8. Rethink create and attach APIs of shared TidStore.

  9. Fix inconsistent function prototypes with function definitions.

  10. Fix a calculation in TidStoreCreate().

  11. Fix potential integer handling issue in radixtree.h.

  12. Add TIDStore, to store sets of TIDs (ItemPointerData) efficiently.

  13. Fix link error for test_radixtree module on Windows

  14. Blind attempt to fix ODR violations

  15. Fix incorrect format specifier for int64

  16. Fix redefinition of typedefs

  17. Add template for adaptive radix tree

  18. Fix signedness error in 9f225e992 for gcc

  19. Introduce helper SIMD functions for small byte arrays

  20. Optimize vacuuming of relations with no indexes.

  21. Add bound check before bsearch() for performance

  22. Allocate consecutive blocks during parallel seqscans