[c-nsp] Replacement for HSRP

[Scott Granados]

This might be like fly swatting with sledge hammers but how about BGP in 
this situation?

Depending on your network addressing particulars and your ISP you could do 
something like have each router set up an IGP session between them over a 
common gig interface, then one gig out of each router to your ISP and then 
tune as needed.  If memory is a concern simply have defaults sent to you and 
should one link fail the route should withdraw automatically.  (as well as 
your announcement)  If you're speaking to the same ISP on both loops you 
could use a private AS.  I might be way way off here but to me this 
shouldn't require any custom scripting or hack work arounds but I may not be 
understanding your setup fully.

Hope this helps.


----- Original Message ----- 
From: "Bulgaria Online - Assen Totin" <assen at online.bg>
To: <cisco-nsp at puck.nether.net>
Sent: Thursday, December 06, 2007 8:15 AM
Subject: [c-nsp] Replacement for HSRP


> Hi all,
>
> I have a pretty weird situation and I'm trying to figure out a
> work-around for it.
>
> We have 2 Cisco 2821 each equipped with a 16-port switch (Service
> Module). Both act as gateways to a local network - each LAN device is
> connected to both Service Modules (bonding interface) and one of the two
> Gi ports of the each router are connected to one and same ISP.
>
> The Service Modules utilize HSRP to provide all LAN devices with a
> default gateway address.
>
> However, the ISP blocks multicast packets on our external interfaces, so
> the routers cannot talk to each other, hence HSRP is not actually
> running - both external interfaces claim to be master because no slave
> can be reached. (The ISP is in Japan and it is difficult to understand
> why they do so; they have not been very co-operative.)
>
> Since the Service Modules are separate entities running their own IOS,
> the router has no way of tracking the status of the internal interface.
>
> My last resort may be to connect the two routers to each other using the
> available Gi ports on each chassis, then somehow monitor this link (if
> one of the devices goes down, the link will go down; this is much worse
> than HSRP, but is still better than nothing) and enable/disable the ISP
> interface, depending on the status of this link (or may be just bring up
> and down a secondary IP address). Basically, this means the ability to
> enable/disable one Ethernet interface depending on the link status of
> another Ethernet interface.
>
> Any ideas how to achieve this (if possible at all) are welcome - up to
> writing a TCL job to run in the router...
>
> Thanks in advance,
>
> Assen Totin
>
>
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/