On Wed, Oct 6, 2010 at 15:34, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Magnus Hagander <magnus@hagander.net> writes:
>> On Wed, Oct 6, 2010 at 15:16, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> However, the usage in pgstat.c is hard-wired, meaning that if you
>>> have a configuration where "localhost" doesn't resolve correctly
>>> for whatever reason, there's no simple recourse to get the stats
>>> collector working. So ISTM there is an argument for changing that.
>
>> Well, hardcoding it will break the (unusual) case when localhost isn't
>> 127.0.0.1 / ::1. (You'd obviously have to have it try both ipv4 and
>> ipv6).
>
> You didn't read what I wrote before. Those numeric addresses define the
> loopback address, *not* "localhost". When localhost fails to resolve
> as those address(es), it's localhost that is wrong. We have actually
> seen this in the field with bogus DNS providers.
>
>> It's not common, but i've certainly come across a number of virtual
>> machines where localhost resolves (through /etc/hosts) to the machines
>> "real" IP rather than 127.0.01, because 127.0.0.1 simply doesn't
>> exist.
>
> That appears to me to be a broken (non RFC compliant) VM setup.
Can't argue with that. But it exists.
> However, maybe what this is telling us is we need to expose the setting?
> Or perhaps better, try 127.0.0.1, ::1, localhost, in that order.
That was kind of my point, that yes, we probably need to do one of
those at least. Today it is "kind of exposed", because you can edit
/etc/hosts - you don't need to rely on DNS for it. I just don't want
to lose that ability.
--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/