| |||||||||||||||||||||||||||||||
|
> At 4:30 PM +0200 5/9/08, Ralf Hildebrandt wrote: > [...] > >A named/BIND9 restart/stop/start did always invoke: > >echo "nameserver 127.0.0.1" | /sbin/resolvconf -a lo.named > >and this caused postfix to be signalled by resolvconf (after all, the > >resolv.conf changed!) > > This wanders back into Postfix relevance perhaps... > > Is there really some reason that one would want to restart/reload > Postfix arbitrarily on all resolv.conf changes? It seems to me that > this would be a judgment call, since so many of the things one > *might* do to resolv.conf would not cause trouble for whatever > persistent or semi-persistent DNS state Postfix might be retaining. On a client box, maybe. On a production server, definitely not. Postfix network processes are by design short-lived so they pickup resolv.conf and other changes on the fly. Unless the machine is only up for a few minutes at a time, Postfix will push out the mail eventually anyway. > I don't oppose allowing admins to be lazy when it is feasible, but > this strikes me as a case where sparing root from engaging in a small > amount of thought could be somewhat expensive. I think this is a result of peecee mentality. Wietse
| ||||||||||||||||||||||||||||||
© 2004-2008 readlist.com