-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Oliver Boehmer (oboehmer) wrote: > > >> >> >>> Oliver, forget what I said... I've read "will prefer" instead of "will >>> never prefer" :-( >>> >>> It's good to know that another provider is using this kind of >>> architecture. It's not something we want to use for all our customers >>> but this specific customer has some constraints >>> which require to loadbalance their traffic. >>> >>> I guess we could also use OSPF and have the same cost for the path CE1 >>> --> PE1 and the second path CE1 --> CE2 --> PE2. What would be the best >>> in this case ? eBGP multihop or OSPF >>> with costs ? >>> >> >> if OSPF is an option, I would prefer this as it is "cleaner" and more >> "natural" routing, and no risk of running into loops. > > and with OSPF you could even come up with an EEM script on the CEs to > adjust the CE->PE link metric based on the HSRP status, so you can even > provide load-sharing when the HSRP master fails over to the "standby" (if > such a failure scenario would still allow the standby CE to perform > load-sharing, that is).. >
Isn't this a classic case for PfR? - -- ========= bep -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.17 (MingW32) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iEYEARECAAYFAlK5Lo8ACgkQE1XcgMgrtyZQIgCfeLlW2xCJS2KBT2SUnaETsbB9 6fYAmwXRRSA+0iVDu/LbPvgrKIEFeFRB =o6be -----END PGP SIGNATURE----- _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
