Thanks, Sasha. Is my understanding for the all-active case correct?
It should be noted that in the scenario described by Chalapathi, only PE1 advertises the MAC/IP route for MAC1. PE2 and PE3 advertise only alias labels.. Regards, Muthu On Thu, Mar 7, 2019 at 3:40 PM Alexander Vainshtein < [email protected]> wrote: > > > Muthu and all, > > Quoting from Section 14.1.1 “Single-Active Redundancy Mode” of RFC 7432: > > > > If the primary PE encounters a failure, it MAY withdraw its set of > > Ethernet A-D per ES routes for the affected ES prior to withdrawing > > its set of MAC/IP Advertisement routes. > > > > If there is only one backup PE for a given ES, the remote PE MAY use > > the primary PE's withdrawal of its set of Ethernet A-D per ES routes > > as a trigger to update its forwarding entries, for the associated MAC > > addresses, to point towards the backup PE. As the backup PE starts > > learning the MAC addresses over its attached ES, it will start > > sending MAC/IP Advertisement routes while the failed PE withdraws its > > routes. This mechanism minimizes the flooding of traffic during > > fail-over events. > > > > If there is more than one backup PE for a given ES, the remote PE > > MUST use the primary PE's withdrawal of its set of Ethernet A-D per > > ES routes as a trigger to start flooding traffic for the associated > > MAC addresses (as long as flooding of unknown unicast packets is > > administratively allowed), as it is not possible to select a single > > backup PE. > > > > So there are actually three sub-cases in the single-active redundancy mode > use case: > > 1. The single-active multi-homed ES has been advertised by exactly > two PEs. In this case withdrawal of the per-ES Ethernet A-D route by the > primary PE may result in other PEs sending the unicast traffic for MAC > addresses learned from this ES to the secondary PE using the alias labels > advertised for the corresponding EVI in the per-EVI Ethernet A-D routes. > > 2. The single-active multi-homed ES has been advertised by three > or more PEs, and flooding of unknown unicast is allowed. In this case > withdrawal of the per-ES Ethernet A-D route by the primary PE MUST result > in flooding of the unicast traffic with unlearned MAC addresses using the > common scheme for BUM traffic. The Aliasing labels are not relevant for > this use case. > > 3. The single-active multi-homed ES has been advertised by three > or more PEs, and flooding of unknown unicast is not allowed. In this case > withdrawal of the per-ES Ethernet A-D route by the primary PE MUST in just > local flooding of the unicast traffic with unlearned MAC addresses. > > > > Hope this helps. > > My 2c, > > Sasha > > > > Office: +972-39266302 > > Cell: +972-549266302 > > Email: [email protected] > > > > *From:* BESS <[email protected]> *On Behalf Of *Muthu Arul Mozhi > Perumal > *Sent:* Thursday, March 7, 2019 8:53 AM > *To:* Luc Andre Burdet (lburdet) <[email protected]> > *Cc:* chalapathi andhe <[email protected]>; Sean Wu < > [email protected]>; Jaikumar Somasundaram < > [email protected]>; [email protected]; > [email protected] > *Subject:* Re: [bess] FW: Regarding the Alias label usage in EVPN > Multihoming > > > > My understanding: > > > > For the single-active case in the diagram below, when PE4 receives the A-D > per ES route withdraw it won't know whether PE2 or PE3 will become the new > primary/DF for the <ES, VLAN>, so it will start flooding the traffic > destined to MAC1. Then either PE2 or PE3 will become the new primary/DF for > the <ES, VLAN> and advertise the MAC/IP route for MAC1. PE4 will then start > sending the traffic destined to MAC1 to the new primary. > > > > For the all-active case, when PE4 receives the A-D per ES route withdraw > it will update the nexthop list for MAC1 by removing PE1 from that list. > PE4 will then load balance the traffic destined to MAC1 by send it to PE2 > and PE3 using alias label. > > > > Regards, > > Muthu > > > > On Wed, Mar 6, 2019 at 7:51 PM Luc Andre Burdet (lburdet) < > [email protected]> wrote: > > Hi Chalu, > > > > Please read 7432 s.8.4: in single-active it is not an aliasing > label/procedure but a backup-path procedure. > > It will answer your questions (both, actually). > > > > There is no flooding once the MAC is learnt & distributed: cf. that’s the > point. > > > > > > [image: > http://www.cisco.com/c/dam/m/en_us/signaturetool/images/banners/standard/09_standard_graphic.png] > > *Luc André Burdet* > > [email protected] > > Tel: *+1 613 254 4814* > > Cisco Systems Canada Co. / Les Systemes Cisco Canada CIE > > Cisco.com <http://www.cisco.com/web/CA/> > > > > > > *From: *BESS <[email protected]> on behalf of chalapathi andhe < > [email protected]> > *Date: *Wednesday, March 6, 2019 at 04:15 > *To: *"[email protected]" <[email protected]> > *Cc: *Sean Wu <[email protected]>, Jaikumar Somasundaram < > [email protected]>, "[email protected]" < > [email protected]> > *Subject: *Re: [bess] FW: Regarding the Alias label usage in EVPN > Multihoming > > > > > > > > Hi All, > > > > Can you please help us on the following issue. > > In the following diagram, PE1, PE2, PE3 are connected to the same ES [ES1] > in Single active mode, and PE4 is a remote PE. > > Let’s say PE1 is advertising the MAC1 route, PE4 will install the MAC1 > with the PE1 as primary path with MAC Label, > > and PE2, PE3 as backup with the Alias Label. Now if the PE1 to CE1 link > goes down, then PE1 withdraw the EAD/ES route > > which will be processed by PE4. > > Now what should the forwarding state at PE4 ?, PE4 should update the > forwarding state of MAC1 with the PE2, PE3 Alias label > > and any traffic destined to MAC1 should be flooded to PE2, PE3 with the > Alias labels ? > > Or packet should be flooded to PE2, PE3 with the Flood labels ? Or should > it be some other method ? > > > > In similar if the ES is operating in all active mode, what should be the > forwarding state at PE4 ? > > PE4 should update the forwarding state of MAC1 with the PE2, PE3 Alias > label > > and any traffic destined to MAC1 should be sent to either PE2 or PE3 with > the Alias labels [ not flood to both] ? > > Or packet should be flooded to PE2, PE3 with the Flood labels or Alias > Label ? > > Or should it be some other method ? > > > > > > [image: cid:1695247dcc04ce8e91] > > > > > > Thanks, > > Chalapathi. > > _______________________________________________ > BESS mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/bess > > > ___________________________________________________________________________ > > This e-mail message is intended for the recipient only and contains > information which is > CONFIDENTIAL and which may be proprietary to ECI Telecom. If you have > received this > transmission in error, please inform us by e-mail, phone or fax, and then > delete the original > and all copies thereof. > ___________________________________________________________________________ >
_______________________________________________ BESS mailing list [email protected] https://www.ietf.org/mailman/listinfo/bess
