Обсуждение: BUG #6110: pg_hba.conf samehost does not work

The following bug has been logged online:

Bug reference:      6110
Logged by:          Binod Nepal
Email address:      nepalbinod@yahoo.com
PostgreSQL version: 9.0.2
Operating system:   Windows 7 Ultimate 32/64
Description:        pg_hba.conf samehost does not work
Details:

pg_hba.conf

# IPv4 local connections:
host    all             all             127.0.0.1/32            md5
host    all             all             samehost            md5
host    all             all             samenet            md5
# IPv6 local connections:
host    all             all             ::1/128                 md5
host    all             all             samehost                md5
host    all             all             samenet                md5


postgresql.conf

listen_addresses = '*'


SYSTEM DETAILS

Operating System
Windows 7 Ultimate 32 or 64 bit

Computer Name
SERVER

PostgreSQL Server Version
PostgreSQL 9.0.2


BUG DESCRIPTION

I am using pgAdminIII, trying create a "New Server Registration". When I
connect to (localhost), it works fine. But when I change the host to the
actual workstation name, it says:

FATAL: no pg_hba.conf entry for host "fe80::c81c:9e9c:6570:d0bf%20", user
"postgres", database "postgres", SSL off

even when I have

host    all             all             samehost                md5

on pg_hba.conf on IPv4 and IPv6 settings.



PLEASE NOTE
I am trying to connect to the server from the same machine.


ADDITIONAL INFORMATION
If I connect to "SERVER" from other computers on the same network, it
works.

For further information, please contact.

On Mon, Jul 11, 2011 at 5:13 AM, Binod Nepal <nepalbinod@yahoo.com> wrote:
> # IPv4 local connections:
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 127.0=
.0.1/32 =A0 =A0 =A0 =A0 =A0 =A0md5
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 sameh=
ost =A0 =A0 =A0 =A0 =A0 =A0md5
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 samen=
et =A0 =A0 =A0 =A0 =A0 =A0md5
> # IPv6 local connections:
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 ::1/1=
28 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 md5
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 sameh=
ost =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0md5
> host =A0 =A0all =A0 =A0 =A0 =A0 =A0 =A0 all =A0 =A0 =A0 =A0 =A0 =A0 samen=
et =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0md5

I don't think it accomplishes anything at all to have the same lines
in there twice.  The comment doesn't affect whether the line applies
to IPv4 or IPv6.

ISTM the problem here must be the Windows version of
pg_foreach_ifaddr(), which uses this method to try to get a list of
addresses:

WSAIoctl(sock, SIO_GET_INTERFACE_LIST, 0, 0,
                                         ii, n_ii * sizeof(INTERFACE_INFO),
                                         &length, 0, 0)

Perhaps there's a better method we could be using, but I have no idea
what it is.  A quick Google search didn't turn up any IPv6-related
restrictions on SIO_GET_INTERFACE_LIST.

--=20
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company