pushed dns servers not being used

[Ralf Hildebrandt]

After the upgrade from yenial to yakkety, the internal DNS servers (pushed by the openvpn server) are not being used by dnsmasq.

Symptom:

I connect via openvpn (called from network manager), the connection is established successfully but INTERNAL hostnames cannot be resolved.

The system itself is using dnsmasq.

I then have to kill the running dnsmasq instance. After it's been (automagically) restarted, internal DNS names can be resolved.

ProblemType: Bug
DistroRelease: Ubuntu 16.10
Package: network-manager-openvpn 1.1.93-1ubuntu1
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: Unity:Unity7
Date: Thu Oct 6 20:06:33 2016
InstallationDate: Installed on 2014-02-07 (972 days ago)
InstallationMedia: Ubuntu 13.10 "Saucy Salamander" - Release amd64 (20131016.1)
SourcePackage: network-manager-openvpn
UpgradeStatus: Upgraded to yakkety on 2016-10-01 (5 days ago)

[Launchpad Janitor (janitor)]

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

[Matthew Gregg (mcg)]

Not sure if I have the same issue, but I have hard set DNS servers in the OpenVPN connection. On first connect it works and I can resolve internal hosts, but if I reconnect the VPN connection, I can no longer resolve until I restart the dnsmasq process. It will restart and I can resolve hosts again.

[Jeroen Hoek (mail-jeroenhoek)]

This issue is now present in 16.04 as well. As Matthew mentions, the first time you connect to a VPN with OpenVPN DNS works, the second time it doesn't.

In case anyone stumbles upon this issue, this workaround (restarting dnsmasq) makes name resolution work again for a single VPN session:

sudo pkill dnsmasq

You can execute it either before or after the VPN connection is made.

[Matthew Gregg (mcg)]

FWIW this seems fixed in 17.04, but a few new issues have been introduced :-)