Hi David,
Thanks for the reply. Your explanation about the multiple connections
actually explains things nicely and fits my observations. will go the
.pgpass route instead.
On 2020-07-19 7:25 p.m., David Christensen wrote:
On Jul 18, 2020, at 6:40 PM, Computerisms Corporation <[email protected]>
wrote:
Hi Folks,
so I get the impression that bucardo is really meant to run in a password only
authentication scheme, but I haven't read any where that peer/ident won't work,
and they are listed on the old wiki pages as options.
I get that I need passwords for remote connections, but I would much prefer if
local connections could use peer authentication instead of md5, mostly because
I intend to try scripting things for future usage and would avoid more
passwords than necessary involved in said script.
Hi Bob,
You should be able to adjust how your local PostgreSQL cluster authenticates
(md5 vs peer vs trust, etc) via the pg_hba.conf.
The main thing that is throwing you for a loop currently is there are actually
two separate types of connections that are made here: ones from the `bucardo`
command-line script, which are made from the user account you run with (so
often the “bucardo” user), and ones made from the plperl functions inside the
database, which will run with the permissions of the “postgres” user.
I recommend instead of trying to bypass the auth type just using a .pgpass file
to store the credentials for the bucardo user, and including that in both
~postgres and ~bucardo (or whatever user is running the `bucardo` control
script.
HTH,
David
--
David Christensen
Senior Software and Database Engineer
End Point Corporation
[email protected]
785-727-1171
--
David Christensen
Senior Software and Database Engineer
End Point Corporation
[email protected]
785-727-1171
_______________________________________________
Bucardo-general mailing list
[email protected]
https://bucardo.org/mailman/listinfo/bucardo-general
