Re: systemd.resolved problems
Jeremy Ardley wrote:
>
> On 25/3/22 7:26 am, Greg Wooledge wrote:
> > On Fri, Mar 25, 2022 at 06:51:55AM +0800, Jeremy Ardley wrote:
> > > Is there any compelling reason to use systemd.resolved over ordinary DNS ?
> > > If not, why was it inflicted on debian?
> > It's disabled by default. It's there if you wish to try it, but out of
> > the box, it does absolutely nothing except sit there taking up space.
> >
>
> Are you sure it's disabled by default? I don't recall converting over to it
> on my various machines. And when I search on it there are lots of pages
> about how to disable it, and virtually none on how to enable it.
>
> I thought there might be some voodoo reason to do with something called dbus
> - of which I know nothing, nor the obscure journald.
>
> Anyway, bind9 works pretty well as a local caching nameserver
It does, and general practice on mailservers used to be that
they got local caching nameservers because they do so much name
resolution.
For smaller networks, I think just having a local caching nameserver
that all the hosts can rely on is sufficient.
unbound is also an excellent caching nameserver, but I wouldn't
bother recommending that you switch unless bind makes you
unhappy in some way.
-dsr-
Reply to: