After reboot the machine can not resolve anything outside the local net and the nets connected via VPN. No DNS.
Other clients (including TW updated some days ago) are doing fine on the network.
Changed from Networkmanager to Wicked and back including reboots.
Changed cables, switches, routers, even complete network.
Nothing helped. Needless to say no update works as well as the servers cannot be resolved.
Now I booted this machine from a TW on an USB-stick and everything works as expected (I’m writing this from the USB-TW booted…).
DNS is in my network ONLY from the router (DNS Resolver with DNSsec via TCP/UDP port 53 to router IP), no way to get out for ANY machine to ANY other DNS-server. Works well for years. Not going to change this.
What has changed recently that could explain the unexpected behaviour regarding DNS?
Something must have corrupted your /etc/resolv.conf file, it defines what nameservers your system queries. I’ve seen NetworkManager do this on occasion, I guess it’s a bug or some other unforeseen consequence.
netupdate simply moved the existing one into a backup and added the right nameservers from your network configuration (in this case, dhcp)
Edit:
/etc/resolv.conf is simply a text file with the follow structure:
search is not necessary but the nameserver defines what IP it queries, for example 8.8.8.8. Each nameserver is put on its own line so if you want to have Google’s DNS servers, it would look like this:
…actually had a look into resolv.conf at some point in time during the DNS blackout, it was completely empty (besides the usual comments). Thought it might no longer be used and left it untouched.
