DNS resolver stops accepting queries

Bug #1629620 reported by Joe Barnett
34
This bug affects 6 people
Affects Status Importance Assigned to Milestone
network-manager (Ubuntu)
Confirmed
Undecided
Unassigned

Bug Description

After a while / after resume from suspend, the system becomes unable to respond to DNS queries, with the following output for example:

$ nslookup google.com 127.0.1.1
Server: 127.0.1.1
Address: 127.0.1.1#53

** server can't find google.com: REFUSED

This is on an up to date yakkety install. Changing resolv.conf to use 8.8.8.8, or restarting network-manager appears to fix this, at least until it recurs again.

ProblemType: Bug
DistroRelease: Ubuntu 16.10
Package: network-manager 1.2.4-0ubuntu1
ProcVersionSignature: Ubuntu 4.8.0-17.19-generic 4.8.0-rc7
Uname: Linux 4.8.0-17-generic x86_64
ApportVersion: 2.20.3-0ubuntu7
Architecture: amd64
CurrentDesktop: GNOME
Date: Sat Oct 1 21:28:51 2016
EcryptfsInUse: Yes
IfupdownConfig:
 # interfaces(5) file used by ifup(8) and ifdown(8)
 auto lo
 iface lo inet loopback
InstallationDate: Installed on 2015-08-21 (407 days ago)
InstallationMedia: Ubuntu-GNOME 15.10 "Wily Werewolf" - Alpha amd64 (20150819)
IpRoute:
 default via 192.168.1.1 dev wlp3s0 proto static metric 600
 169.254.0.0/16 dev docker0 scope link metric 1000 linkdown
 172.17.0.0/16 dev docker0 proto kernel scope link src 172.17.0.1 linkdown
 192.168.1.0/24 dev wlp3s0 proto kernel scope link src 192.168.1.5 metric 600
NetworkManager.state:
 [main]
 NetworkingEnabled=true
 WirelessEnabled=true
 WWANEnabled=true
 WimaxEnabled=true
ProcEnviron:
 TERM=xterm-256color
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: network-manager
UpgradeStatus: Upgraded to yakkety on 2016-09-29 (2 days ago)
nmcli-dev:
 DEVICE TYPE STATE DBUS-PATH CONNECTION CON-UUID CON-PATH
 docker0 bridge connected /org/freedesktop/NetworkManager/Devices/1 docker0 63f0b05b-f699-4a68-81ce-f62d2c1a247d /org/freedesktop/NetworkManager/ActiveConnection/0
 wlp3s0 wifi connected /org/freedesktop/NetworkManager/Devices/4 HOME-1622-5 6df3f2f0-fb86-43b5-802a-d59bc5734be3 /org/freedesktop/NetworkManager/ActiveConnection/2
 24:DA:9B:B9:09:0E bt disconnected /org/freedesktop/NetworkManager/Devices/3 -- -- --
 lo loopback unmanaged /org/freedesktop/NetworkManager/Devices/2 -- -- --
nmcli-nm:
 RUNNING VERSION STATE STARTUP CONNECTIVITY NETWORKING WIFI-HW WIFI WWAN-HW WWAN
 running 1.2.4 connected started full enabled enabled enabled enabled enabled

Revision history for this message
Joe Barnett (thejoe) wrote :
Revision history for this message
Joe Barnett (thejoe) wrote :

this seems to be less of a problem after switching /etc/resolv.conf to point to the systemd-resolved managed file instead of the resolvconf managed file. not sure exactly which package should be changed to have this work automatically?

Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in network-manager (Ubuntu):
status: New → Confirmed
Revision history for this message
Joshua Anderson (joshua-anderson) wrote :

I noticed this issue on one of the two computers I upgraded from xenial to yakkety. Specifically the affected computer was an 2016 Dell XPS 13. They're both similarly configured ubuntu installations with Broadcom nics, so it's curious that this only affects one of them. I'm not sure what debug information would be helpful, but I'm happy to provide anything that would be useful.

Revision history for this message
Richard (ismail-a) wrote :

When the network interface providing dns is managed by network manager, dns fails on relaunch of the interface

test dns:
dig x$(date +%N).blogspot.com +short
if it fails, there is no output

verify that Internet is available:
dig @8.8.8.8 x$(date +%N).blogspot.com +short
blogspot.l.googleusercontent.com.
216.58.194.129

provoke the problem:
- unplug and re-plug the network interface (usb Ethernet)
- or resume from standby

quick-fix to regain working dns: restart Network Manager
sudo systemctl restart NetworkManager
- dns now works again

Cause: for Ubuntu 16.10 Yakketi, this appeared shortly before 160926
probably with network-manager package 1.2.2-0ubuntu8

It appears the problem is fixed permanently by installing an old version of network-manager
wget https://launchpad.net/ubuntu/+source/network-manager/1.2.2-0ubuntu4/+build/10026930/+files/network-manager_1.2.2-0ubuntu4_amd64.deb
sudo dpkg -i network-manager_1.2.2-0ubuntu4_amd64.deb
- this completes for 16.10 on 161112

Revision history for this message
Richard (ismail-a) wrote :

dns appears to be ok, it's just not working
for example, one can do:
nmcli d show | grep DNS
IP4.DNS[1]: 192.168.1.15

the Network Manager-provided dnsmasq is just not using that dns server that network manager knows about.

Revision history for this message
Richard (ismail-a) wrote :
Revision history for this message
Harald Rudell (harald-rudell) wrote :

This is a dup of #1639776

A patched deb package is here: https://bugs.launchpad.net/ubuntu/+source/dnsmasq/+bug/1639776/comments/7

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.