From: Antonin Houska <ah@cybertec.at>
To: Mihail Nikalayeu <mihailnikalayeu@gmail.com>
cc: Alvaro Herrera <alvherre@alvh.no-ip.org>,
    Fujii Masao <masao.fujii@gmail.com>, Robert Treat <rob@xzilla.net>,
    Pg Hackers <pgsql-hackers@lists.postgresql.org>
Subject: Re: Adding REPACK [concurrently]
In-reply-to: 
 <CADzfLwWdw=5KhXE5Fr3khytp4mW2girFVJHHb4jSgEncc6237Q@mail.gmail.com>
References: 
 <CADzfLwXx46j8KwQjjM1ZcqNBsx-k6GxHOzDJkm4SHjh+cv31Rw@mail.gmail.com>
 <202508091333.qvgvo7ikuezm@alvherre.pgsql>
 <CADzfLwW=b=U3e6aasi=XorN8hZSiCKZErKs9qhyK7m=w=wokAg@mail.gmail.com>
 <40729.1755799624@localhost>
 <CADzfLwXpyzGGVB+nbSAMBWDBhzTyn6F2hRrAqcyJd3c6gT2EOQ@mail.gmail.com>
 <9536.1756127358@localhost>
 <CADzfLwWdw=5KhXE5Fr3khytp4mW2girFVJHHb4jSgEncc6237Q@mail.gmail.com>
Comments: In-reply-to Mihail Nikalayeu <mihailnikalayeu@gmail.com>
   message dated "Mon, 25 Aug 2025 16:15:16 +0200."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <21930.1756136535.1@localhost>
Content-Transfer-Encoding: quoted-printable
Date: Mon, 25 Aug 2025 17:42:15 +0200
Message-ID: <21931.1756136535@localhost>
Archived-At: 
 <https://www.postgresql.org/message-id/21931.1756136535%40localhost>
Precedence: bulk

Mihail Nikalayeu <mihailnikalayeu@gmail.com> wrote:

> > Not sure I understand in all details, but I don't think SnapshotSelf i=
s the
> > correct snapshot. Note that HeapTupleSatisfiesSelf() does not use its
> > 'snapshot' argument at all. Instead, it considers the set of running
> > transactions as it is at the time the function is called.
> =

> Yes, and it is almost the same behavior when a typical MVCC snapshot
> encounters a tuple created by its own transaction.
> =

> So, how it works in the non MVCC-safe case (current patch behaviour):
> =

> 1) we have a whole initial table snapshot with all the xmin =3D repack X=
ID
> 2) appling transaction sees ALL the self-alive (no xmax) tuples in it
> because all tuples created\deleted by transaction itself
> 3) each update/delete during the replay selects the last existing
> tuple version, updates it xmax and inserts a new one
> 4) so, there is no any real MVCC involved - just find the latest
> version and create a new version
> 5) and it works correctly because all ordering issues were resolved by
> locking mechanisms on the original table or by reordering buffer

ok

> How it maps to MVCC-safe case (SnapshotSelf):
> =

> 1) we have a whole initial table snapshot with all xmin copied from
> the original table. All such xmin are committed.
> 2) appling transaction sees ALL the self-alive (no xmax) tuple in it
> because its xmin\xmax is committed and SnapshotSelf is happy with it

How does HeapTupleSatisfiesSelf() recognize the status of any XID w/o usin=
g a
snapshot? Do you mean by checking the commit log (TransactionIdDidCommit) =
?

> 3) each update/delete during the replay selects the last existing
> tuple version, updates it xmax=3Doriginal xid and inserts a new one
> keeping with xmin=3Dorignal xid
> 4) --//--
> 5) --//--

> > However, at the time we're replaying the UPDATE in the new table, the =
tuple
> > may have been already deleted from the old table, and the deleting tra=
nsaction
> > may already have committed. In such a case, HeapTupleSatisfiesSelf() w=
ill
> > conclude the old version invisible and the we'll fail to replay the UP=
DATE.
> =

> No, it will see it - because its xmax will be empty in the repacked
> version of the table.

You're right, it'll be empty in the new table.

-- =

Antonin Houska
Web: https://www.cybertec-postgresql.com