[1050] in linux-net channel archive
Re: IPIP tunneling: a practical example?
daemon@ATHENA.MIT.EDU (John Paul Morrison)
Tue Sep 5 15:46:23 1995
From: jmorriso@bogomips.ee.ubc.ca (John Paul Morrison)
To: iialan@iifeak.swan.ac.uk (Alan Cox)
Date: Mon, 4 Sep 1995 19:55:40 -0800 (PDT)
Cc: Lutz.Pressler@AMS.Med.Uni-Goettingen.DE, linux-net@vger.rutgers.edu
In-Reply-To: <m0spY97-00013gC@iiit.swan.ac.uk> from "Alan Cox" at Sep 4, 95 10:48:52 am
>
> > I think. I'm afraid I wasn't able to get it done with Linux 1.3.22/23
> > on both sides (and net-tools-1.3.6 BETA3). Maybe I just fundamentally
> > misuderstood something.. or there are still some bugs :)
>
> I am not confident that 1.3.2x IPIP works. Its on my current test and clean
> up list as well as making the tunnel driver run at a sensible speed (needed
> for multicast output). Try with 1.3.3 and IPIP and if that works suspect
> a bug with 1.3.2x
I tested IPIP in 1.3.23 and it works for in/outgoing TCP stuff. Outgoing
pings working, but it still won't reply to pings because it sends out
corrupted datagrams, so routers drop them. 1.3.23 is at the level of 1.3.8
or so (works but has bugs) for IPIP. IPIP stopped working briefly.
I don't know why IPIP is so tricky to get working. It doesn't bode
well for other services that use encapsulation (mobile IP,
compression, encryption, and multiprotocol support: tunneling IPX over
IP, IP over AppleTalk or whatever).
I keep looking at the tunnel driver and wonder what's wrong. It
certainly looks like it's recalculationg the IP header checksum
correctly.
>
> Alan
>
>
---------------------------------------------------------------------------
BogoMIPS Research Labs -- bogosity research & simulation -- VE7JPM --
jmorriso@bogomips.ee.ubc.ca ve7jpm@ve7jpm.ampr.org jmorriso@ve7ubc.ampr.org
---------------------------------------------------------------------------
