Re: [bess] A question about RFC 8317

2019-01-26 Thread Aldrin Isaac
Btw, the other problem with “two RTs” scheme might be mobility. The leaf MAC-VRF can’t see each others sequence numbers for a MAC. Please see/consider my prior email. Need to address E-Tree for non-MPLS encaps and in DC settings (think PVLAN). Cheers, Aldrin On Thu, Dec 20, 2018 at 11:42 PM

Re: [bess] A question about RFC 8317

2019-01-26 Thread Aldrin Isaac
Hi guys, If we instead consider that the two-RTs scheme from operator point-of-view is really a Root-only and Leaf-only MAC-VRF scheme (vs mixed root/leaf MAC-VRF). Suppose all routes from a leaf VRF are marked with a “leaf indication”. Suppose we use only one RT where leaf MAC-VRF rejects any

Re: [bess] A question about RFC 8317

2018-12-20 Thread Alexander Vainshtein
Jorge, Lots of thanks for a prompt response. My conclusion js tbat the "two RTs" scheme should be used with special care in E-tree solutions. This was not my impression from the first reading of 8317. Since the "two RTs" scheme is very popular in hub-and-spoke" solutiobs for IP VPN, the fact

Re: [bess] A question about RFC 8317

2018-12-20 Thread Alexander Vainshtein
Ali, Lots of thanks for a prompt response. I fully agree that tbe use case I've described can be addressed by the genegal techniques of RFC 8317. I only wanted to u dersrand applicability of fbe "two RTs" scheme, and both Jorge and you confifm that it would result in undesirable behavior in

Re: [bess] A question about RFC 8317

2018-12-20 Thread Alexander Vainshtein
ain View); Alexander Vainshtein; Ali Sajassi (sajassi) Cc: John E Drake (jdr...@juniper.net); Samer Salam (ssalam); ju1...@att.com; sbout...@vmware.com; bess@ietf.org Subject: Re: [bess] A question about RFC 8317 Sasha, I would add only to Jorge’s response that in your topology below: “PE3 would floo

Re: [bess] A question about RFC 8317

2018-12-20 Thread Ali Sajassi (sajassi)
Hi Sasha, The use case that you have described below is a legitimate use case and if we look at what happens in RFC 7432 baseline, there is no flooding there because MAC addresses among multi-homing PEs get synch’d up and thus even a PE in the all-active multi-homing group doesn’t receive a

Re: [bess] A question about RFC 8317

2018-12-20 Thread Rabadan, Jorge (Nokia - US/Mountain View)
rake (jdr...@juniper.net)" , "Samer Salam (ssalam)" , "ju1...@att.com" , "sbout...@vmware.com" , "bess@ietf.org" Subject: Re: [bess] A question about RFC 8317 Sasha, I would add only to Jorge’s response that in your topology below: “PE3 would flood anyt

Re: [bess] A question about RFC 8317

2018-12-20 Thread Luc Andre Burdet (lburdet)
tein , "Ali Sajassi (sajassi)" Cc: "John E Drake (jdr...@juniper.net)" , "Samer Salam (ssalam)" , "ju1...@att.com" , "sbout...@vmware.com" , "bess@ietf.org" Subject: Re: [bess] A question about RFC 8317 Hi Sasha, What you are explai