Public bug reported:
After upgrade to Ubuntu 18.10, my connection to my employer's VPN server
has become unstable.
The server provides both ipv4 and ipv6 endpoints. Since I have global
ipv6 at home, my client defaults to connecting over ipv6.
The connection consistently drops after only a few minutes of being
connected.
If I force the connection to use ipv4 (by hard-coding the host
resolution in /etc/hosts), the connection is stable.
If I use ipv6 and force proto-tcp=yes, the connection is stable.
If I use the default ipv6 and udp, the connection is not stable.
Using tcp for VPNs is suboptimal. Hard-coding addresses in /etc/hosts
is suboptimal.
This may be a bug in openvpn rather than in network-manager-openvpn, but
I'm starting here.
It's possible there are pathMTU issues involved, but I reproduced this
same problem when I was on a different network from my home network
(though I did not confirm, when I had access to this other network, that
it was ipv6-enabled).
I will work on getting more details about both sides of the network to
try to debug this.
** Affects: network-manager-openvpn (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Desktop
Packages, which is subscribed to network-manager-openvpn in Ubuntu.
https://bugs.launchpad.net/bugs/1800542
Title:
OpenVPN connection not stable after upgrade to 18.10 (udp, ipv6)
Status in network-manager-openvpn package in Ubuntu:
New
Bug description:
After upgrade to Ubuntu 18.10, my connection to my employer's VPN
server has become unstable.
The server provides both ipv4 and ipv6 endpoints. Since I have global
ipv6 at home, my client defaults to connecting over ipv6.
The connection consistently drops after only a few minutes of being
connected.
If I force the connection to use ipv4 (by hard-coding the host
resolution in /etc/hosts), the connection is stable.
If I use ipv6 and force proto-tcp=yes, the connection is stable.
If I use the default ipv6 and udp, the connection is not stable.
Using tcp for VPNs is suboptimal. Hard-coding addresses in /etc/hosts
is suboptimal.
This may be a bug in openvpn rather than in network-manager-openvpn,
but I'm starting here.
It's possible there are pathMTU issues involved, but I reproduced this
same problem when I was on a different network from my home network
(though I did not confirm, when I had access to this other network,
that it was ipv6-enabled).
I will work on getting more details about both sides of the network to
try to debug this.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/network-manager-openvpn/+bug/1800542/+subscriptions
--
Mailing list: https://launchpad.net/~desktop-packages
Post to : [email protected]
Unsubscribe : https://launchpad.net/~desktop-packages
More help : https://help.launchpad.net/ListHelp
