Horton

Re: Add 64-bit XIDs into PostgreSQL 15

Pavel Borisov <pashkin.elfe@gmail.com>

View thread
From: Pavel Borisov <pashkin.elfe@gmail.com>
To: Maxim Orlov <orlovmg@gmail.com>
Cc: Aleksander Alekseev <aleksander@timescale.com>, Andres Freund <andres@anarazel.de>, Ilya Anfimov <ilan@tzirechnoy.com>, Postgres hackers <pgsql-hackers@lists.postgresql.org>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2022-03-02T14:43:11Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Add SLRU tests for 64-bit page case

  2. Make use FullTransactionId in 2PC filenames

  3. Use larger segment file names for pg_notify

  4. Index SLRUs by 64-bit integers rather than by 32-bit integers

Hi hackers!

Hi! Here is the rebased version.
>
I'd like to add a description of what was done in v9:
- The patch is rebased on current master branch
- In-memory tuple storage format was refactored as promised to have
pre-calculated 64bit xmin and xmax, not just copies of pd_xid_base and
pd_multi_base.
- Fixed bug reported by Andres Freund, with lazy conversion of pages
upgraded from 32 to 64 xid when first tuple read (and therefore lazy
conversion) is done in read-only state (read-only xact or on replica). In
this case now in memory buffer descriptor will be marked
with REGBUF_CONVERTED flag. When cluster comes to read-write state this
will lead to emitting full page write xlog instruction for this page.

Relevant changes in README are also done.

We'd very much appreciate enthusiasm to have 64 bit xid's in PG15 and any
effort to review and test this feature.

Alexander, thanks for your attention to the patchset. Your questions and
review are very much welcome!
The participation of other hackers is highly appreciated as always!

--
Best regards,
Pavel Borisov

Postgres Professional: http://postgrespro.com <http://www.postgrespro.com>