Simplest way I know of is to write a C function which fires off a
shell script which in turn executes psql with appropriate parameters.
The C function can pass it individual variables, or the entire SQL
string.
The reason for the shell script is that way the database, port, etc.
are easy to change and don't require a re-compile / reload.
One thing you definately cannot do is a C function which creates a new
database connection on it's own. Doesn't like that at all!
I've implemented a form of 'su' for postgres doing that -- makes a
second connection back into the database. A middle man shell script
ensures that only a few specific requests can be made (one is actually
creating a system user with very specific parameters).
--
Rod Taylor
There are always four sides to every story: your side, their side, the
truth, and what really happened.
----- Original Message -----
From: "Marc Wrubleski" <mlwruble@sorexsoftware.com>
To: <pgsql-general@postgresql.org>
Sent: Friday, April 06, 2001 7:22 PM
Subject: [GENERAL] update tables in remote db using trigger
> Hi all,
>
> I would like to do something like the following in a trigger:
>
> update tablename on otherdbhost set field1 = new.field1 where
keyfield =
> new.keyfield;
>
> Is this a pipe dream using Postgres or is there some way to perform
such
> a feat?
>
> Thanks.
>
> Marc Wrubleski
>
>
> ---------------------------(end of
broadcast)---------------------------
> TIP 4: Don't 'kill -9' the postmaster
>