Re: [HACKERS] expand_dbname in postgres_fdw

Поиск
Список
Период
Сортировка
От Ashutosh Bapat
Тема Re: [HACKERS] expand_dbname in postgres_fdw
Дата
Msg-id CAFjFpRejPgSVdzwvx5M-qV36nN6Sgtz5zDDPx5z+KLZ6BK+QkQ@mail.gmail.com
обсуждение исходный текст
Ответ на Re: [HACKERS] expand_dbname in postgres_fdw  (Tom Lane <tgl@sss.pgh.pa.us>)
Ответы Re: [HACKERS] expand_dbname in postgres_fdw  (Arseny Sher <a.sher@postgrespro.ru>)
Список pgsql-hackers
On Thu, Jul 27, 2017 at 12:21 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> On Wed, Jul 26, 2017 at 5:38 AM, Ashutosh Bapat
>> <ashutosh.bapat@enterprisedb.com> wrote:
>>> According to F.34.1.1 at [1] passing connection string as dbname
>>> option should work, so your question is valid. I am not aware of any
>>> discussion around this on hackers.
>
>> I kind of wonder if this had some security aspect to it?  But not sure.
>
> The main problem to my mind is that a connection string could possibly
> override items meant to be specified elsewhere.  In particular it ought
> not be allowed to specify the remote username or password, because those
> are supposed to come from the user mapping object not the server object.
> I suspect you could break things by trying to specify client_encoding
> there, as well.

+1.

>
> In any case, I entirely reject the argument that the existing
> documentation says this should work.  It says that you can specify (most
> of) the same fields that are allowed in a connection string, not that one
> of those fields might be taken to *be* a connection string.
>

Section F.34.1.1. at [1] says
"A foreign server using the postgres_fdw foreign data wrapper can have
the same options that libpq accepts in connection strings, as
described in Section 33.1.2, except that these options are not
allowed:". When it says, " accepts same options", users would
interpret it as "accept in the same manner as specified in the
referenced section". Also, dbname is not one of the listed exceptions,
so a user would expect same behaviour when the same value for dbname
option is provided in foreign server options and libpq connection
string. In the referenced section "dbname" is described as

--
dbname

The database name. Defaults to be the same as the user name. In
certain contexts, the value is checked for extended formats; see
Section 33.1.1 for more details on those.
--

There is some grey area where different people will interpret those
sentences in different manner. So, may be better to say that "dbname"
option in foreign server accepts only database names.

[1] https://www.postgresql.org/docs/10/static/postgres-fdw.html#idm44880567492496
-- 
Best Wishes,
Ashutosh Bapat
EnterpriseDB Corporation
The Postgres Database Company



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Chapman Flack
Дата:
Сообщение: Re: [HACKERS] AdvanceXLInsertBuffer vs. WAL segment compressibility
Следующее
От: Masahiko Sawada
Дата:
Сообщение: Re: [HACKERS] Transactions involving multiple postgres foreign servers