Re: NM gets it wrong



On Fri, 2005-11-18 at 00:45 +0100, Tim Niemueller wrote:
> Not the inet addr and P-t-p addresses, they are the same! I have
> double-checked. The OpenVPN service sends the correct information to NM.
> I added output to the openvpn service and used dbus-monitor to see what
> actually gets send.
> I'm not even sure if this is a problem. A route to the remote address is
> correctly set. In the example given this looks like:
> tim evilgenius:[~/dev]# route -n
> Kernel IP Routentabelle
> Ziel            Router          Genmask         Flags Metric Ref    Use
> Iface
> 10.9.0.5        192.168.4.1     255.255.255.255 UGH   0      0        0 ath0
> 192.168.4.0     0.0.0.0         255.255.255.0   U     0      0        0 ath0
> 169.254.0.0     0.0.0.0         255.255.0.0     U     0      0        0 lo
> 0.0.0.0         0.0.0.0         0.0.0.0         U     0      0        0 tun0
> 
> And although I cannot ping the remote IP I'm still able to route traffic
> over that connection.
> 
> Does anybody have an idea where this gets messed up and why and where NM
> sets the P-t-P address or what I did so horribly wrong and actuall _if_
> this is a problem. Did you do any more tests j?

Right, this is all an artifact of the vpnc implementation.  In vpnc, the
PtP and local address are the same.  This was something only a second
VPN implementation could find :)

The NMIP4Config structure doesn't have an entry for PtP addresses.
We'll likely need to add one and also augment the vpn-service -> NM API
to deal with that, so that each individual vpn service can do what it
wants to with that.  Which variable is the "server" address that gets
passed from OpenVPN to your helper?  Is it VPNGATEWAY?

> Another thing that I noticed that if you edit two VPN connections of
> different VPN type (vpnc and openvpn) one after another the second time
> it will show you both widgets stacked on one screen. Is that a problem
> that I introduced or something that is not been seen yet since there has
> been only one VPN service till now?

We should probably bug davidz about this one...

Dan




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]