[nycbug-talk] Odd log messages with OpenVPN bridge

Jonathan jonathan
Thu Feb 3 16:15:23 EST 2005


Tillman Hodgson wrote:
> On Thu, Feb 03, 2005 at 03:21:48PM -0500, Jonathan wrote:
> 
>>I keep getting the following messages in my daily security run and 
>>clearly they have something to do with network looping but I don't know 
>>what causes them or how to look into them further.  I did try but Google 
>>turned up nothing at all :P
>>
>>server.kc8onw.net kernel log messages:
>>
>>
>>>>-- loop (0) 00.40.05.49.45.12 to rl0 from tap0 (active)
>>>>-- loop (1) 00.40.05.49.45.12 to tap0 from rl0 (active)
>>>>-- loop (2) 00.40.05.49.45.12 to rl0 from tap0 (active)
>>>>-- loop (3) 00.40.05.49.45.12 to tap0 from rl0 (active)
>>>>-- loop (4) 00.40.05.49.45.12 to rl0 from tap0 (active)
>>>>-- loop (5) 00.40.05.49.45.12 to tap0 from rl0 (active)
>>>>-- loop (6) 00.40.05.49.45.12 to rl0 from tap0 (active)
>>>>-- loop (7) 00.40.05.49.45.12 to tap0 from rl0 (active)
>>>>-- loop (8) 00.40.05.49.45.12 to rl0 from tap0 (active)
>>>>-- loop (9) 00.40.05.49.45.12 to tap0 from rl0 (active)
>>>>-- loop (10) 00.40.05.49.45.12 to rl0 from tap0 (active)
> 
> 
> Just out of curiousity, do the errors go away if you use tun devices
> instead of tap devices in your OpenVPN config?
> 
> My thinking is as follows: Tap devices are a "raw" transport. You
> generally only need them if you want to carry non-IP traffic (IPX comes
> to mind). Tun devices are "IP" devices, so we can eliminate an entire
> layer of the stack by using them (as well as making OpenVPN more
> efficient -- less overhead per packet means more data per data).
> 
> -T
> 
I don't really know, if it comes down to that I may try it but the 
reason for using tap in the first place was for things like IPX and 
games that depend on UDP broadcasts :P

Thanks,
Jonathan




More information about the talk mailing list