Project

General

Profile

Actions

Bug #8075

closed

OpenVPN binds to wrong interface with no ip on first interface

Added by Adrian Zaugg about 7 years ago. Updated over 5 years ago.

Status:
Rejected
Priority:
Normal
Assignee:
-
Category:
OpenVPN
Target version:
-
Start date:
11/09/2017
Due date:
% Done:

0%

Estimated time:
Plus Target Version:
Release Notes:
Affected Version:
2.4.1
Affected Architecture:

Description

With two WAN interfaces and with an OpenVPN server on each, bound to its
interface, there is a wrong IP assertion in case the first interfaces
does not get an IP.

In Detail:
- a system with two WAN interfaces that both get their IP by DHCP
- on each WAN there is an OpenVPN Server configured, bound to one iface
If for some reason no IP can be obtained on the first WAN interface, the
OpenVPN instance actually bound to WAN1 does grab the IP of the WAN2
interface. The second OpenVPN instance cannot start then and mourns
"address already in use".

I would expect that the OpenVPN instance of WAN1 doesn't grab the
address of WAN2.

Actions #1

Updated by robi robi about 7 years ago

Workaround: bind both OpenVPNs to 127.0.0.1 (localhost) (use different ports). Then forward these ports to their respective WANs (can use the same port numbers on the different WANs).

Actions #2

Updated by Jim Pingle over 5 years ago

  • Category set to OpenVPN
  • Status changed from New to Rejected

The workaround is the proper solution here if you need to use the same port on multiple WANs.

Actions

Also available in: Atom PDF