"repliation" as database name
Hackorum builds and tests every patch posted to the lists, not only commitfest submissions. This is Hackorum's own CI rather than the PostgreSQL project's, and it is still under testing - please report anything that looks wrong.
You can run a PostgreSQL built from this patch straight from Docker, with no checkout and no build:
docker run --rm -p 5432:5432 ghcr.io/hackorum-dev/postgres-patch:t39851psql -h localhost -U postgresBuilt from patchset v2 (message #2), July 27, 2026 at 07:47 PM.
Every patchset is also pushed to a branch of our PostgreSQL fork, so you can check out the same tree CI built. Without a PostgreSQL checkout:
git clone --branch t39851_2 https://github.com/hackorum-dev/postgres.gitIn a checkout you already have, add the fork once:
git remote add hackorum https://github.com/hackorum-dev/postgres.gitthen, for this patchset and every later one:
git fetch hackorum t39851_2 && git checkout t39851_2Patchset v2 (message #2) is on t39851_2
Hello.
We can create a database named "replication".
$ createdb replication
A pg_hba.conf entry with DATABASE="all" is described as 'does not
match "replication"' in the comment there, but actually it
matches and we can connect to the database
"replication". (Documentation doesn't mention the restriction)
$ psql replication -At -c 'select current_database()'
replication
We can specify the name replication by quoting and it does not
match a replication connection. It is not documented at all.
pg_hba.conf
local "replication" all trust
#local replication all trust ## commented out
FATAL: could not connect to the primary server: FATAL: no pg_hba.conf entry for replication connection from host "[local]", user "horiguti", SSL off
$ psql replication -At -c 'select current_database()'
replication
The same can be said to sameuser, samerole and even all. I think
this is absolutely sane behavior and worth documentation in any
extent if it doesn't become complex.
I think that at least the following amendments would be needed.
- Remove ""all" does not match "replication"". Instead "The "all"
keyword does not match replication connections."
- double-quoted database name is taken literally.
Is it worth doing?
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
At Tue, 18 Dec 2018 18:48:25 +0900 (Tokyo Standard Time), Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> wrote in <20181218.184825.02619975.horiguchi.kyotaro@lab.ntt.co.jp>
- Remove ""all" does not match "replication"". Instead "The "all"
keyword does not match replication connections."
- double-quoted database name is taken literally.
I found that in the documentation thanks to a notification
off-list. And after some reconfirmation, what I want to fix is
only a few lines of comment in pg_hba.conf.sample.
-# database name, or a comma-separated list thereof. The "all"
-# keyword does not match "replication". Access to replication
-# must be enabled in a separate record (see example below).
+# database name, or a comma-separated list thereof. The "all" keyword
+# matches all databases. The "replication" keyword matches a physical
+# replication connection request and it must be enabled in a separate
+# record (see example below)regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> writes:
I found that in the documentation thanks to a notification
off-list. And after some reconfirmation, what I want to fix is
only a few lines of comment in pg_hba.conf.sample.
-# database name, or a comma-separated list thereof. The "all" -# keyword does not match "replication". Access to replication -# must be enabled in a separate record (see example below). +# database name, or a comma-separated list thereof. The "all" keyword +# matches all databases. The "replication" keyword matches a physical +# replication connection request and it must be enabled in a separate +# record (see example below)
Hm, I agree that the para doesn't read very well now, but I think this
could be improved further. How about something like
# DATABASE can be "all", "sameuser", "samerole", "replication", a
# database name, or a comma-separated list thereof. The "replication"
# keyword matches replication connection requests (see example below).
# The "all" keyword matches all database names, but not replication
# connections.
regards, tom lane
At Wed, 26 Dec 2018 12:59:32 -0500, Tom Lane <tgl@sss.pgh.pa.us> wrote in <32289.1545847172@sss.pgh.pa.us>
Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> writes:
I found that in the documentation thanks to a notification
off-list. And after some reconfirmation, what I want to fix is
only a few lines of comment in pg_hba.conf.sample.-# database name, or a comma-separated list thereof. The "all" -# keyword does not match "replication". Access to replication -# must be enabled in a separate record (see example below). +# database name, or a comma-separated list thereof. The "all" keyword +# matches all databases. The "replication" keyword matches a physical +# replication connection request and it must be enabled in a separate +# record (see example below)Hm, I agree that the para doesn't read very well now, but I think this
could be improved further. How about something like# DATABASE can be "all", "sameuser", "samerole", "replication", a
# database name, or a comma-separated list thereof. The "replication"
# keyword matches replication connection requests (see example below).
# The "all" keyword matches all database names, but not replication
# connections.
I'm afraid that just dropping "it must be enabled in a separate
record" leads to confusion. How about adding a comment to
replication connection examples.
# Allow replication connections from localhost, by a user with the
# replication privilege. Each definition must have its own record.
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
At Mon, 28 Jan 2019 17:30:57 +0900 (Tokyo Standard Time), Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp> wrote in <20190128.173057.41178374.horiguchi.kyotaro@lab.ntt.co.jp>
At Wed, 26 Dec 2018 12:59:32 -0500, Tom Lane <tgl@sss.pgh.pa.us> wrote in <32289.1545847172@sss.pgh.pa.us>
Hm, I agree that the para doesn't read very well now, but I think this
could be improved further. How about something like# DATABASE can be "all", "sameuser", "samerole", "replication", a
# database name, or a comma-separated list thereof. The "replication"
# keyword matches replication connection requests (see example below).
# The "all" keyword matches all database names, but not replication
# connections.I'm afraid that just dropping "it must be enabled in a separate
record" leads to confusion. How about adding a comment to
replication connection examples.# Allow replication connections from localhost, by a user with the
# replication privilege. Each definition must have its own record.
Mmm, this doesn't seem to saying what I wanted to say there.
This seems better.
# Allow replication connections from localhost, by a user with
# the replication privilege. They must have separate records from
# non-replication connections.
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center