DNS fails a few seconds after restarting NetworkManager

English Network/Internet
#1

Hi, I recently got a weird problem with my remote tumbleweed server.

  1. I can ssh to it.
  2. after restarting NetworkManager through systemd, everything works fine, but only for a few seconds.(ping a domain resolves alright but stops work after about 8 packets.)
  3. after problems in 2, I cannot even ping the dns server (the outgoing traffic seems all stopped). but the ssh connection stays well alive.
  4. in stage 1-3, /etc/resolve.conf updates normally and is correctly set.

I’ve tried resetting iptables, checking firewall rules, but everything seems normal but the problem presists, any ideas? Thanks in advance.

#2

Do you ping the DNS server based on name or IP address? If it is IP address, can you try tracepath instead?

Do you have a firewall running? If so can you (temporary) disable it and check if that fixed the problem?

#3

Tried both, same outcome (ping fails a few seconds after NetworkManager restarts).

Whether I can ping through it or not, tracepath only gets responses from localhost, as follows (test IP corresponds to bing.com):


$ tracepath 13.107.21.200
 1?: [LOCALHOST]                      pmtu 1500
 1:  no reply
 2:  no reply
 3:  no reply
 4:  no reply
 5:  no reply
 6:  no reply

Tried the above tests with firewall stopped, same result.

#4

Can you start on the command line “nmcli monitor” and restart NetworkManager? What output do you get?

What is “ip a” and “sudo ethtool <active-if?” showing when the interface is down?

#5

It’s a remote server, I’m afraid I won’t be able to test with the interface down. I’ll report back when I get access on-site. Thanks for the help.