[c-nsp] mBGP problem

[Michael Robson]

> trying to remember the transition stuff cisco supplied.  
> can't you set the nlri in an in-bound route-map?  seem to 
> recall that there was a neighbor setting that caused routes 
> to get both nlri - translate-update, but that may have just 
> been circa ~11.3.
>
After a long delay...

I have tried adding the line

neighbor <ipv4-address> translate-update ipv4 multicast unicast

to the ipv4 unicast family for the core router which peers with this mBGP
unaware router but this doesn't seem to be showing the expected routes when
I issue

sh bgp ipv4 multicast

Route-maps don't seem to have a "set nlri..." option on our 12.2(18)SXF4
routers.

Anyone any ideas?

Ta,

Michael.

 
> Sun, May 27, 2007 at 10:36:14AM +0100, Michael Robson:
> > We have a mBGP core routers(Cisco 6500, sup720s) that peer 
> separately 
> > for multicast and unicast routes (using address families) 
> and a site 
> > that, at present, doesn't support these BGP address families. The 
> > problem is that, because this site router doesn't support this, the 
> > multicast BGP table isn't being populated for the site's 
> addresses on 
> > the core and an RPF check is failing further upstream. We have a 
> > cludgy fix in place involving statics that doesn't really 
> scale and so 
> > I was looking for a better temporary solution until this 
> site router 
> > has been upgraded. Is there any way of getting the unicast 
> BGP routes to populate the multicast BGP table?
> > 
> > Michael.
> > 
> > _______________________________________________
> > 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/
>