On Wed, Jan 05, 2011 at 09:54:53PM +0100, Paweł Krześniak wrote:
2011/1/4 Daniel P. Berrange <berrange(a)redhat.com>:
> NB, there was a discussion with the dnsmsaq maintainer a few
> months back now about changing the dnsmasq architecture such
> that we only need one dnsmasq process.
is this consistent with Laine's arguments about radvd?
read last paragraph of this:
https://www.redhat.com/archives/libvir-list/2010-December/msg00912.html
What Laine writes is simply our current thinking about radvd.
Since usage of radvd is an internal implementation detail, we
are free to change our mind about this at any time in the future.
Likewise for dnsmasq. Exposing the functionality you propose in
the XML restricts our flexibility in this respect
Daniel