Yes, if your upstream's routers are overloaded, you'll still be blackholed until they can process the change, no matter how fast you change.
----- Mike Hammett Intelligent Computing Solutions Midwest Internet Exchange The Brothers WISP ----- Original Message ----- From: "Trey Scarborough" <t...@3dsc.co> To: af@af.afmug.com Sent: Wednesday, April 19, 2023 4:55:30 PM Subject: Re: [AFMUG] BGP Question You can prevent this by having a backup default route. The other thing is if you are n a physical interface or aggregated interface with a vlan on your upstream's router. If its physical it should stop routing traffic when it sees the interface down. If not it may keep trying to send traffic down that interface. BFD can help with this. It will Also depend on how your provider handles the traffic when the the session with you terminates. If it will forward it on to your other providers or not. If your alternate providers are connected closely to you it shouldn't take to long, but it can take 10-20min for the loss of the peer to propagate across all global routing tables. On 4/19/23 9:47 AM, Mark - Myakka Technologies wrote: > We have two circuits coming into our NOC. We peer with 3 different providers > on each circuit. If a circuit fails, BGP will do it's magic and traffic will > start flowing though the surviving circuit. > > However, we seemed to get about 5 - 10 minutes of unstable Internet while > this is happening. Is this normal? If not, what can I do to speed up the > process? Is it a function of my routers having to rebuild routes? Will new > faster routers help? Is it a function of timers? Keep-Alive is 30s and hold > is at 90s. Should I investigate BFD? > > How fast could I expect to get this fail-over to work under best conditions? > > Thanks, > Mark mailto:m...@mailmt.com > > Myakka Communications > www.Myakka.com > > -- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com
-- AF mailing list AF@af.afmug.com http://af.afmug.com/mailman/listinfo/af_af.afmug.com