[Openvpn-users] Odd tap driver dhcp issue

[Kris Jordan <openvpn.kj@xxxxxxxxxx>]

Before posting all my configuration details and such, I'm going to try
to keep things simple...

Searching the archives I see that plenty of people have DHCP problems,
but all I've seen so far have to do with it basically not working at
all. My setup seems to work perfectly at first glance, but then it falters.

Linux on the server side, Centos3. Windows 2k/XP on the client side. I'm
using the latest stable OpenVPN version on both ends.

I'm using tap on both sides. And I'm using the built-in DHCP server.
When I click "Reconnect" in OpenVPN GUI (running OpenVPN directly and
clicking F3 gives me the same problem) everything works good like on the
first connect, but soon after, the routes that OpenVPN has added, disappear.

I noticed this error in the event log along with it saying it will
assign a 169 address, though it quickly gets the correct address, DNS,
etc. before I can see that it's using that address.

------------------------------------------------------------------------
Event Type:	Warning
Event Source:	Dhcp
Event Category:	None
Event ID:	1003
Date:		12/4/2006
Time:		12:15:49 AM
User:		N/A
Computer:	X
Description:
Your computer was not able to renew its address from the network (from
the DHCP Server) for the Network Card with network address 00FFA3985FD9.
  The following error occurred:
The semaphore timeout period has expired. . Your computer will continue
to try and obtain an address on its own from the network address (DHCP)
server.

For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.
Data:
0000: 79 00 00 00               y...
------------------------------------------------------------------------

This happens for 2k and XP.

So it seems to do something like this:

1. Connects
2. Sets DHCP information provided by tap driver.
3. Gets DHCP information again and fails.
4. Sets it to 169.*
5. Loses routes.
6. Gets DHCP information again and succeeds.

This all happens within 30seconds.

On XP, this also happens a lot on the first connection.

Adding "route-delay 30" solves the problem as the routes are added after
this DHCP problem happens, but I hate this delay.

This problem doesn't seem to happen using tun. But I would like it to
work with the existing OpenVPN setup (old multiple port design + DHCPD).
All connected clients are added to a bridge, br0. DHCPD listens on br0
so that it doesn't have to listen to tap+ and be restarted for each
client connection.

I use server-bridge with the new single port design config, but I still
add it to br0 so that I don't have to reconfigure the firewall. Turning
off DHCPD doesn't help.

I also tried 2.1rc1, but it worked similarly.

I tried tap-debug driver, I do see plenty of back and forth DHCP messages.

ipconfig /renew also doesn't seem to work good either on an already
connected tunnel. /release then /renew works fine though.

Thanks a lot!

- Kris Jordan -

______________________
OpenVPN mailing lists
https://lists.sourceforge.net/lists/listinfo/openvpn-users