09-19-2011 04:31 PM - edited 03-04-2019 01:39 PM
We currently have an existing 3945 doing BGP with our ISP. We'd like to put in a second router and utilise HSRP on our internal side to improve redundancy. Our ISP has suggested we do separate BGP peering sessions on the 2 routers towards them and use AS Pre-ending to favour a particular path back into our AS. I've attached a diag.
I'f we do this, how do we influence the internet to re-route back into us if one of the internal interfaces say Router A F0/0 goes down? Or should we also run HSRP on the ISP facing interfaces?
Appreciate suggestions.
Thks
Terry
09-19-2011 06:43 PM
pre-pending will take make an AS path to a peer longer which will be less preferred. so if you pre-pend on one router on the WAN peer that will make it less preferred. your two routers should have iBGP between them. i do not think you need HSRP but it may depend on what you want to achieve. lets hear what the others have to say.
09-19-2011 06:59 PM
hi there
if the internal switch is L3 capable then run iBGP between the switch and the two routers with out using HSRP
advertise the internal networks from the internal switch using iBGP with both routers if Fa0/0 fail then Router1 will stop advertising and Router2 will be used even if you do AS Prepending
however now because you are going to change the active standby method from HSRP to pure ibgp/eBGP routing
then you need to let the internal switch prefer the route coming from router1 for outbound direction this is can be done by using local preference
in router 1 increase the local preference for the route coming from the ISP side and apply this route-map in the inbound direction from the ISP Peer
this way R1 qill be the main path for outbound and inbound traffic and the other path is a failover
the other method is to Keep the HSRP and use static routes to point to the internal switch in each of the routers for the internal networks point to the internal switch relevant interface and use ip sla/ tracking with those static routes to monitor the interface status of fa0/0
redistribute this static route into BGP and if the interface is down i mean fa0/0 the static routes will go down and the route will be withdrawn from the BGP routing table at this router
but if you switch support L3 routing using first method simpler and more scalable
Hope this help
09-19-2011 07:05 PM
personally i would go L3. once configured correctly it will work flawlessly.
09-19-2011 07:22 PM
cool,
yes the Switch is L3 Capable so will try out the iBGP methodolgy...looks much simpler.
Thks
Terry
09-20-2011 12:10 AM
I've tested it out and it works. Switchover takes approx 180 sec due to the holdtimers of BGP. If I were to modify this what would be a suggested value? 10sec Hello 30sec Holddown. Are there any side effects to lowering these values?
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: