[prev in list] [next in list] [prev in thread] [next in thread] 

List:       zebra
Subject:    [zebra 20114] RE: stopping RIP adv.of lan routes when lan cable d
From:       Eric Polin <eric.polin () netwolves ! com>
Date:       2003-07-31 21:28:22
[Download RAW message or body]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

BTW --

This happens with ospf as well. I have found that when the internal
interface is faulty, i can run a vrrp off script that will kill ospf/zeb,
and the reverse routes will work properly with the secondary (backup) box,
and then when the interface becomes active again, i can run zebra/ospf(or
rip), and the routes recover. This method works, however it is deffinatly
not clean. i would rather have an internal command in zeb/ospf/rip to
determine that the lan network is down (or the lan card has no
carrier/down/etc..), and when that network is unreachable, it should stop
advertising the route. When the network is reachable again, it should
advertise. 

thansk for any help.

Eric Polin
System Engineer


- -----Original Message-----
From: Eric Polin [mailto:eric.polin@netwolves.com]
Sent: Tuesday, July 29, 2003 9:09 AM
To: 'zebra@zebra.org'
Subject: [zebra 19965] stopping RIP adv.of lan routes when lan cable
disconnected


- -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- - --REPOST WITH PROPER SUB--

Hello zebra- 

I am a long time lurker, this is my first post. I am having a problem with
RIP. My topology includes the following setup: ZEBRA/RIPD, VRRP, GRE, RIP,
RACOON, with all running on FREEBSD-4.8-STABLE. I have /32bit vpn tunnels
established from a dual-box vrrp solution to the HQ(single box), and i pass
GRE through that tunnel for my rip routes. I am about 95% done with the
solution, and can emulate a failure by pulling the external links from the
MASTER (active) vrrp box, and then through RIP it will fail back, when it
recieves the RIP updates through the GRE connection. 

I have a slight problem, i am emulating a failure on the LAN/SWITCH port, by
pulling the internal cable, when i do this, all traffic gets routed through
to the active vrrp (which is the second box, with a higher metric), and then
continues to the other side of the vpn tunnel. However, the reverse path is
still showing the lower metric, since the external connection is still
advertising. (btw, i advertise the same network for both boxes in the vrrp
solution) Is there a way to prevent advertising on the first box, if it
cannot see its internal network?

If needed, i can cat the ripd/zebra conf files.

Any help is appreciated.
thanks in advance,

Eric Polin
Systems Engineer I
- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (MingW32)
Comment: For info see http://www.gnupg.org

iEYEARECAAYFAj8mcqIACgkQIgLfkGwqBEGUCACg8xeke5aTknGnO3aVNTNBvgOc
RXwAn3Ve4Q4e1RjsLGIz1pbBJYydrY/M
=M85C
- -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (MingW32)
Comment: For info see http://www.gnupg.org

iEYEARECAAYFAj8pitQACgkQIgLfkGwqBEG+ZQCgvu0/sT/H64rFV6uFIoDFm4nN
SIMAoLgv4ALWeEJj4bkEHnS4DKKdS0gS
=yVM1
-----END PGP SIGNATURE-----
[prev in list] [next in list] [prev in thread] [next in thread]