[mpls-linux-general] Cisco traceroute issues

[Ben G. <mpl...@my...>]

I have a test network setup using two Cisco 2610 routers running IOS
version 12.2(8)T5.  These routers are configured as PE routers with a CE
router attached to each 2610.  The PE routers have a MPLS-VPN configured
with MP-BGP and can successfully pass traffic from the CE to the other CE.
I have verified that the label stacks are functioning correctly. I have a
linux box running the latest public release version of mpls-linux, ldp,
and zebra 0.93a.  The linux box is acting as a P router.

The lab setup looks something like this:

             CE1 -- PE1 -- P -- PE2 -- CE2


Everything works fine in this setup until I run a traceroute on the CE
towards the other CE.  When I do this, the linux box will lock hard.
There are no crash messages, no kernel panics, just completely frozen.  I
managed to trace the issue to the following command on the cisco routers:
mpls ip propagate-ttl.  When I disable that command, the traceroute works
fine.  I can reproduce this problem at will.  Is there any further
information you all would like to see in regards to this issue?

Thanks,
-Ben