From: Tom Lane <tgl@sss.pgh.pa.us>
To: Adnan Dautovic <daut@mailbox.org>
cc: Adrian Klaver <adrian.klaver@aklaver.com>, pgsql-general@postgresql.org
Subject: Re: Failure of postgres_fdw because of TimeZone setting
In-reply-to: <98F8816F-0AC2-4308-961F-4C7A4FC2C646@mailbox.org>
References: <5DF49366-10D1-42A4-99BF-F9A7DC3AB0F4@mailbox.org> <faf7d602-57b6-47de-87ab-a61baa9ccca1@aklaver.com> <0849E257-0A15-4C2F-A41B-7E0E4B716CAE@mailbox.org> <1273542.1712326418@sss.pgh.pa.us> <98F8816F-0AC2-4308-961F-4C7A4FC2C646@mailbox.org>
Comments: In-reply-to Adnan Dautovic <daut@mailbox.org>
	message dated "Wed, 10 Apr 2024 21:38:52 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <4170839.1712780507.1@sss.pgh.pa.us>
Date: Wed, 10 Apr 2024 16:21:47 -0400
Message-ID: <4170840.1712780507@sss.pgh.pa.us>
Archived-At: <https://www.postgresql.org/message-id/4170840.1712780507%40sss.pgh.pa.us>
Precedence: bulk

Adnan Dautovic <daut@mailbox.org> writes:
> On 05. Apr 2024, at 16:13, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Out of curiosity, does
>> SET timezone to 'GMT';
>> work?

> Yes, it yields:
>> SET
>> 
>> Query returned successfully in 84 msec.

I expected that, because the name "GMT" is hard-wired in our code.
Doesn't help for postgres_fdw though, because it has "UTC" hardwired.
(I have a todo item to rationalize that...)

> By the way, the row count of pg_timezone_names is 385, but I do
> not know how that compares to a more standard installation.

Using current PG HEAD (with tzdata release 2024a):

=# select count(*) from pg_timezone_names;
 count 
-------
   597
(1 row)

I can believe older tzdata releases varying from that a little,
but they haven't exactly been adding zone names at a rapid clip.
Either the one you're dealing with is VERY old or it lost some
files sometime.

			regards, tom lane