> > > Do you have nscd running? If so, turn it off and retry all of your tests. > If it is still misbehaving, then these are the next things I'd look at. > > Is eth1 on the same subnet as the 192.168.9.1 name server? Which interface > has the default route? > > What happens when you do a tcpdump of traffic going out on eth0/eth1 to > port 53? > > What options did you give strace when you were checking with the > functioning resolv.conf and the broken resolv.conf. > > Hugh > > The tcpdump settings is interesting... with resolv.conf set to:
search clean.io nameserver 10.10.10.4 nameserver 192.168.9.1 do standard query A for myserver.example.com on 10.10.10.4 Get response back from 10.10.10.4 do standard query A for myserver.example.com on 192.168.9.1 get failure back Repeat the above but look for myserver.example.com.example.com So it seems the host command cycles through all the dns server entries in /etc/resolv.conf and is somehow not satisfied. Maybe it is because it does not find mx entries... which host command would be looking for. Worth testing I will add MX entries and see if that changes the behaviour. Regards -- Gerhardus Geldenhuis
_______________________________________________ rhelv5-list mailing list rhelv5-list@redhat.com https://www.redhat.com/mailman/listinfo/rhelv5-list
