1

I have recently installed 2 virtual machines using Ubuntu 18.04, both appear to be configured the same, but one will not resolve addresses.

Both servers have a minimal install.

Server 1 runs nginx proxy server. Server 2 runs bind for external DNS

Both servers point to 10.0.181.1 and 10.0.181.2 in their netplan config for DNS.

Both servers have no firewall enabled locally.

Both servers can ping the internal DNS servers.

Server 2: 

root@uksvl-ops01:/etc/netplan# ping uksvl-web01-rs PING uksvl-web01-rs (10.0.52.1) 56(84) bytes of data. 64 bytes from uksvl-web01-rs (10.0.52.1): icmp_seq=1 ttl=64 time=1.19 ms

Server 1:

root@uksvl-web01-ki:/etc/nginx/conf.d# ping uksvl-web01-rs ping: uksvl-web01-rs: Temporary failure in name resolution

I have seen references to disabling systemd-resolve, this has no impact.

There is no DHCP involved here, everything is static.

Improve this question
5
  • 1
    Check the actual content of /etc/resolv.conf and make sure it is sensible. Commented Aug 9, 2018 at 12:54
  • Possible duplicate of Temporary failure in name resolution Commented Aug 9, 2018 at 13:11
  • @MichaelHampton resolv.conf is identical on both. Commented Aug 9, 2018 at 13:40
  • 2
    Huh. I just noticed that you aren't using FQDNs. You should get in the habit of doing so always. Short names are no longer reliable. Commented Aug 9, 2018 at 13:55
  • Did you ever solve this? Could you mark an answer or provide one yourself? Commented Mar 26, 2019 at 11:02

4 Answers 4

Reset to default
1

You probably have a problem connecting to the DNS server.

  • try using dig to resolve the host with your internal DNS: dig @your-internal-dns uksvl-web01-rs
  • check if the DNS-port is open for your virtual host: telnet your-internal-dns 53
Improve this answer
2
  • telnet only tests TCP where DNS uses also UDP. So no need to do this, just use dig +tcp to force TCP connection, or dig +notcp to force UDP. Commented Aug 9, 2018 at 16:35
  • Today I won't recommend telnet anyway. You may want to use netcat (nc) instead, which even supports UDP. @PatrickMevzek is correct: you may just add the +tcp or +notcp option, but nc may also help. Commented May 27, 2022 at 10:28
1

After configuration netplan (on LXC container) with new DNS servers, and after netplan apply, my /etc/systemd/resolved.conf was without DNS servers, so I added them manually and executed service systemd-resolved restart

This is a temporary solution, but I need this only for remove netplan. Maybe this will help others too.

Improve this answer
0

I had very similar issue. I disabled dnssec (DNSSEC=no) in /etc/systemd/resolved.conf and it helped.

Improve this answer
1
  • 2
    It's unlikely that DNSSEC is the reason for problems with private zones. Commented Jan 29, 2019 at 19:50
-3

create a host entry in /etc/hosts. then try

Improve this answer
1
  • 4
    Doesn't really help me solve the issue, I have a perfectly working DNS server that has 100's of entries, adding them all to hosts is not a practical approach. Commented Aug 9, 2018 at 12:32

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.