Invalidate the subscription worker in cases where a user loses their superuser status

From: vignesh C <vignesh21(at)gmail(dot)com>
To: PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Invalidate the subscription worker in cases where a user loses their superuser status
Date: 2023-09-22 18:52:00
Message-ID: CALDaNm2Dxmhq08nr4P6G+24QvdBo_GAVyZ_Q1TcGYK+8NHs9xw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi,

The subscription worker was not getting invalidated when the
subscription owner changed from superuser to non-superuser. Here is a
test case for the same:
Publisher:
CREATE USER repl REPLICATION PASSWORD 'secret';
CREATE TABLE t(i INT);
INSERT INTO t VALUES(1);
GRANT SELECT ON t TO repl;
CREATE PUBLICATION p1 FOR TABLE t;

Subscriber (has a PGPASSFILE for user "repl"):
CREATE USER u1 SUPERUSER;
CREATE TABLE t(i INT);
ALTER TABLE t OWNER TO u1;
-- no password specified
CREATE SUBSCRIPTION s1
CONNECTION 'dbname=postgres host=127.0.0.1 port=5432 user=repl'
PUBLICATION p1;

ALTER USER u1 NOSUPERUSER: -- Change u1 user to non-superuser

Publisher:
INSERT INTO t VALUES(1);

Subscriber:
SELECT COUNT(*) FROM t; -- should have been 1 but is 2, the apply
worker has not exited after changing from superuser to non-superuser.

Fixed this issue by checking if the subscription owner has changed
from superuser to non-superuser in case the pg_authid rows changes.
The attached patch has the changes for the same.
Thanks to Jeff Davis for identifying this issue and reporting it at [1].

[1] - https://www.postgresql.org/message-id/5dff4caf26f45ce224a33a5e18e110b93a351b2f.camel%40j-davis.com

Regards,
Vignesh

Attachment Content-Type Size
v1-0001-Restart-the-apply-worker-if-the-subscription-owne.patch text/x-patch 5.2 KB

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2023-09-22 19:58:55 Re: Questions about the new subscription parameter: password_required
Previous Message Matthias van de Meent 2023-09-22 15:50:08 Re: GenBKI emits useless open;close for catalogs without rows