MIME-Version: 1.0
References: <19439-8f02018f9fc3a240@postgresql.org> <CABPTF7Wx=WTHa67PJUnH8VaJf-svrcFjt1ZgOVrqWzfgwDuMCA@mail.gmail.com>
 <CABPTF7Wb6VOac1EFopGzhAtiHOU8_XRkFvhC6BnSj=jACBB+Gw@mail.gmail.com> <CABPTF7VSNTvHTpS4ga1MfLYO-=xt7mLQ0ONEFobRERnhzXSzxw@mail.gmail.com>
In-Reply-To: <CABPTF7VSNTvHTpS4ga1MfLYO-=xt7mLQ0ONEFobRERnhzXSzxw@mail.gmail.com>
From: Xuneng Zhou <xunengzhou@gmail.com>
Date: Mon, 30 Mar 2026 11:14:19 +0800
Message-ID: <CABPTF7XO101mBvA9u01nHkvt1gaQ3i60hypYOvNp6DdOFnY0jQ@mail.gmail.com>
Subject: Re: BUG #19439: pg_stat_xact_user_tables stat not currect during the transaction
To: klemen.kobau@gmail.com, pgsql-bugs@lists.postgresql.org
Content-Type: multipart/mixed; boundary="000000000000daa566064e353fc1"
Archived-At: <https://www.postgresql.org/message-id/CABPTF7XO101mBvA9u01nHkvt1gaQ3i60hypYOvNp6DdOFnY0jQ%40mail.gmail.com>
Precedence: bulk

--000000000000daa566064e353fc1
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Sun, Mar 29, 2026 at 11:17=E2=80=AFAM Xuneng Zhou <xunengzhou@gmail.com>=
 wrote:
>
> On Sat, Mar 28, 2026 at 1:47=E2=80=AFPM Xuneng Zhou <xunengzhou@gmail.com=
> wrote:
> >
> > On Fri, Mar 27, 2026 at 10:15=E2=80=AFAM Xuneng Zhou <xunengzhou@gmail.=
com> wrote:
> > >
> > > Hi klemen,
> > >
> > > Thanks for the report.
> > >
> > > On Fri, Mar 27, 2026 at 5:36=E2=80=AFAM PG Bug reporting form
> > > <noreply@postgresql.org> wrote:
> > > >
> > > > The following bug has been logged on the website:
> > > >
> > > > Bug reference:      19439
> > > > Logged by:          klemen kobau
> > > > Email address:      klemen.kobau@gmail.com
> > > > PostgreSQL version: 18.0
> > > > Operating system:   Linux (EndeavorOS)
> > > > Description:
> > > >
> > > > I am running postgres:18.0 in a docker container, the configuration=
 is as
> > > > follows:
> > > >
> > > >   postgres:
> > > >     image: postgres:18.0
> > > >     command: [
> > > >       "postgres",
> > > >       "-N", "200",
> > > >       "-c", "max_prepared_transactions=3D100",
> > > >       "-c", "wal_level=3Dlogical",
> > > >       "-c", "max_wal_senders=3D10",
> > > >       "-c", "max_replication_slots=3D20",
> > > >       "-c", "wal_keep_size=3D10",
> > > >       "-c", "max_slot_wal_keep_size=3D1024"
> > > >     ]
> > > >     environment:
> > > >       POSTGRES_USER: postgres
> > > >       POSTGRES_PASSWORD: postgres
> > > >       POSTGRES_DB: postgres
> > > >       TZ: UTC
> > > >       PGTZ: UTC
> > > >     ports:
> > > >       - 5432:5432
> > > >     volumes:
> > > >       - postgres_data:/var/lib/postgresql
> > > >
> > > > I use psql version 18.3.
> > > >
> > > > I run the following:
> > > >
> > > > psql -h localhost -p 5432 -U postgres -d postgres -c "
> > > > -- Transaction 1: insert 1 row, check stats, commit
> > > > BEGIN;
> > > > CREATE TABLE IF NOT EXISTS temp.xact_test (id serial PRIMARY KEY, v=
al text);
> > > > INSERT INTO temp.xact_test (val) VALUES ('a');
> > > > SELECT relname, n_tup_ins FROM pg_stat_xact_user_tables WHERE relna=
me =3D
> > > > 'xact_test';
> > > > COMMIT;
> > > >
> > > > -- Transaction 2: insert 1 row, check stats
> > > > BEGIN;
> > > > INSERT INTO temp.xact_test (val) VALUES ('b');
> > > > SELECT relname, n_tup_ins FROM pg_stat_xact_user_tables WHERE relna=
me =3D
> > > > 'xact_test';
> > > > ROLLBACK;
> > > >
> > > > -- Cleanup
> > > > DROP TABLE temp.xact_test;
> > > > "
> > > >
> > > > And the output is
> > > >
> > > > BEGIN
> > > > CREATE TABLE
> > > > INSERT 0 1
> > > >   relname  | n_tup_ins
> > > > -----------+-----------
> > > >  xact_test |         1
> > > > (1 row)
> > > >
> > > > COMMIT
> > > > BEGIN
> > > > INSERT 0 1
> > > >   relname  | n_tup_ins
> > > > -----------+-----------
> > > >  xact_test |         2
> > > > (1 row)
> > > >
> > > > ROLLBACK
> > > > DROP TABLE
> > >
> > > I can also reproduce this behavior on HEAD.
> > >
> > > > This is not what I would expect from reading
> > > > https://www.postgresql.org/docs/current/monitoring-stats.html#MONIT=
ORING-STATS-VIEWS,
> > > > where it states
> > > >
> > > > > A transaction can also see its own statistics (not yet flushed ou=
t to the
> > > > shared memory statistics) in the views pg_stat_xact_all_tables,
> > > > pg_stat_xact_sys_tables, pg_stat_xact_user_tables, and
> > > > pg_stat_xact_user_functions. These numbers do not act as stated abo=
ve;
> > > > instead they update continuously throughout the transaction.
> > > >
> > > > based on this, I would expect that the numbers are updated each tim=
e an
> > > > insert happens and that they are isolated per transaction.
> > > >
> > >
> > > This snippet of the doc feels somewhat ambiguous. The current behavio=
r
> > > does not seem fully consistent with wording such as =E2=80=9Cthrougho=
ut the
> > > transaction.=E2=80=9D It also seems more appropriate for these table =
views to
> > > reflect stats within their own scopes, rather than cumulative values
> > > spanning txn boundaries.
> > >
> >
> > I=E2=80=99ve looked into this issue and 'd like to propose a patch to a=
ddress
> > it. Feedback is very welcome.
> >
> > --- Root cause
> >
> > Since the stats subsystem was rewritten (commit 5891c7a8), each
> > backend keeps a hash of pending stats entries that persist across
> > transaction boundaries. Entries that flush successfully are deleted
> > from the backend-local pending list; entries not flushed remain
> > pending with their accumulated counters. Flushing is driven by
> > pgstat_report_stat(), called from the ReadyForQuery path, and subject
> > to a rate limiter.
> >
> > The pg_stat_xact_* views read these pending entries directly via
> > find_tabstat_entry() / find_funcstat_entry(). Both functions return
> > the raw accumulated counters without per-transaction scoping.
> >
> > The bug is deterministic when multiple top-level transactions are
> > processed within a single simple-query message, because there is no
> > ReadyForQuery boundary between the transactions and therefore no
> > opportunity to flush and remove the pending entry.
> >
> > --- Fix
> >
> > The patch introduces a per-entry "transaction baseline" =E2=80=94 a laz=
y
> > snapshot of the counters taken the first time an entry is touched in
> > each new top-level transaction. The accessor functions
> > (find_tabstat_entry, find_funcstat_entry) then subtract the baseline
> > from the current counters, yielding only the current transaction's
> > delta.
> >
> > The baseline is keyed by MyProc->vxid.lxid. A static inline helper,
> > pgstat_ensure_xact_baseline(), is called at every nontransactional
> > counter-increment site (the 7 event-counter macros in pgstat.h and the
> > 4 non-inline counter functions in pgstat_relation.c). After the first
> > call per entry per transaction, the check reduces to a single integer
> > comparison with a predictably not-taken branch.
> >
> > For function stats, a new PgStat_FunctionPending wrapper struct embeds
> > PgStat_FunctionCounts at offset 0 with the baseline fields appended,
> > so the flush callback requires only a trivial cast change.
> >
> > --- Testing
> >
> > The regression test is a TAP test under src/bin/psql/t/ that uses psql
> > -c to send multi-statement strings as single simple-query messages.
> > src/bin/psql/t look like the right existing harness for the primary
> > regression because psql -c sends the whole multi-statement string as
> > one simple-query message via simple_action_list dispatch. By contrast,
> > ordinary pg_regress SQL files are executed by psql in file/stdin mode,
> > which runs through MainLoop() and dispatches top-level statements one
> > at a time via SendQuery(), creating ReadyForQuery boundaries that
> > could mask the bug.
> >
> > The test covers three scenarios: table n_tup_ins and seq_scan
> > counters, and function calls =E2=80=94 each verifying isolation across
> > consecutive top-level transactions within a single message.
> >
>
>  postgres % cat output.txt
> BEGIN
> CREATE TABLE
> INSERT 0 1
>   relname  | n_tup_ins
> -----------+-----------
>  xact_test |         1
> (1 row)
>
> COMMIT
> BEGIN
> INSERT 0 1
>   relname  | n_tup_ins
> -----------+-----------
>  xact_test |         2
> (1 row)
>
> ROLLBACK
> DROP TABLE
>
> I can also reproduce this behavior in v13. So this issue pre-existed
> the major refactor 5891c7a8.
>

After further thought of v1, I'd like to also propose an alternative
approach. The root cause and testing rationale are the same as
described upthread; what changes is how the baseline is established.

--- Problems with the lazy (v1) approach
The v1 patch snapshots the baseline lazily -- on every
nontransactional counter increment, it checks whether baseline_lxid
matches the current
transaction and snapshots if not. This has two drawbacks:

1) Hot-path overhead
Every pgstat_count_heap_scan(), pgstat_count_buffer_read_block(), etc.
acquires a branch. After the first call per entry per transaction, the
branch is predictably not-taken, but it could still present on some of
the most frequently executed paths in the backend.

2) Maintenance burden
Every new counter added to PgStat_TableCounts requires a corresponding
baseline snapshot at its increment site. Miss one, and that counter
silently leaks across transactions

--- eager baseline sweep
The attached patch records the baseline eagerly at transaction
boundaries instead of lazily at counter-increment sites.
pgstat_set_pending_baselines() iterates the pgStatPending list and
snapshots each entry's current counts into an xact_baseline field via
struct assignment. It is called from AtEOXact_PgStat() (after folding
transactional counts and removing dropped entries) and from
PostPrepare_PgStat() (after relation cleanup), covering commit, abort,
and PREPARE TRANSACTION. The view accessors unconditionally subtract
the baseline. For entries created in the current transaction,
xact_baseline is zero-initialized, so the subtraction is a no-op.

I don=E2=80=99t have a clear preference between the two approaches; both ar=
e
presented for review.

--
Best,
Xuneng

--000000000000daa566064e353fc1
Content-Type: application/octet-stream; 
	name="v1-0001-Fix-pg_stat_xact_-views-leaking-stats-across-tran.patch"
Content-Disposition: attachment; 
	filename="v1-0001-Fix-pg_stat_xact_-views-leaking-stats-across-tran.patch"
Content-Transfer-Encoding: base64
Content-ID: <f_mnclu85g0>
X-Attachment-Id: f_mnclu85g0
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--000000000000daa566064e353fc1--