On Mon, Mar 14, 2011 at 05:54:58PM -0500, Mark Felder wrote: > On Mon, 14 Mar 2011 14:46:28 -0500, Stuart Henderson > <s...@spacehopper.org> wrote: > > >Make sure your nexthops are valid: bgpctl sh nex > > I worked with my coworker on it this afternoon and he discovered the > nexthops issue. We have resolved the problem for now. > > Out next step is to figure out how to make OpenBGPD be kind of like > a local BGP cache and not actually route traffic through these boxes > because we really don't want that. If we have to we will, but we'd > prefer to not be pushing all our traffic through non-network > hardware if possible. The goal here is to get OpenBGPD to simply be > a better point of contact to the transport routers' BGP than Cisco. > Does anyone know if this type of setup is possible? >
Not exactly sure what you try to achive but it sounds like you plan to setup route-reflectors in your network. Route-reflectors normaly will not see any traffic if you set them up properly. iBGP sessions will not modify nexthops so traffic should flow directly to the edge routers without hitting the reflector. -- :wq Claudio
