Gyan, You're most welcome.
Yours Irrespectively, John Juniper Business Use Only From: Gyan Mishra <hayabusa...@gmail.com> Sent: Sunday, March 22, 2020 5:26 PM To: John E Drake <jdr...@juniper.net> Cc: Rishabh Parekh <risha...@gmail.com>; SPRING WG <spring@ietf.org>; Voyer, Daniel <daniel.vo...@bell.ca> Subject: Re: [spring] SR replication segment for P2MP MDT Thanks John!! Did a quick read and does appear the concept of P2MP tree "P2MP LSM like" instantiation optimization alternative to IR P2P replication which is process intensive on the head end FEC root SR source node. Appears it is backwards compatible to IR if utilized. Great!! Thanks Gyan On Sun, Mar 22, 2020 at 4:59 PM John E Drake <jdr...@juniper.net<mailto:jdr...@juniper.net>> wrote: https://tools.ietf.org/html/draft-shen-spring-p2mp-transport-chain-01<https://urldefense.com/v3/__https:/tools.ietf.org/html/draft-shen-spring-p2mp-transport-chain-01__;!!NEt6yMaO-gk!SZJ9KrZb5FDnLgfvSpVcwP1SN97M1awV1MKU3fNf7mYhEME91N98Z71-Iqb9JzI$> Yours Irrespectively, John Juniper Business Use Only From: spring <spring-boun...@ietf.org<mailto:spring-boun...@ietf.org>> On Behalf Of Gyan Mishra Sent: Sunday, March 22, 2020 4:45 PM To: Rishabh Parekh <risha...@gmail.com<mailto:risha...@gmail.com>> Cc: SPRING WG <spring@ietf.org<mailto:spring@ietf.org>>; Voyer, Daniel <daniel.vo...@bell.ca<mailto:daniel.vo...@bell.ca>> Subject: Re: [spring] SR replication segment for P2MP MDT On Sat, Mar 21, 2020 at 3:22 PM Rishabh Parekh <risha...@gmail.com<mailto:risha...@gmail.com>> wrote: Gyan, >The SR replication segment tree sid draft states that it can only be used for >PCE centralized controller model. >I am guessing BIER maybe an option? Strictly speaking, BIER is not SR dataplane, but yes it is an option. So BIER technically can be used with SR but has its own BIFT forwarding table. Understood. For the WG question I posed is that if LDP is eliminated from SP core and you don't want to use RSVP-TE and BIER is not yet available and SR replication tree SID draft can only be used if a centralized PCE is utilized what alternatives or options does the operators have? Also if PCE is configured on SR source node in a hybrid model and BGP prefix SID is advertise by PCC elements in IGP via BGP LS propagation would that be an option for SR-MPLS multicast P2MP and MP2MP LMDTs? Any other options for operators? LDP with RFC 7473 can be used just for mLDP LSPs. So in the case where you have LDP still enabled in the core and have SR-PREFER configured so all L3 vpn unicast traffic is using SR-MPLS forwarding plane - so then for multicast to work to use mLDP data plane this draft and what you are saying for that to work is you have to configure a knob to disable FEC root prefix lsp. Not sure how that would work as that mldp fec binding for mLDP core p-tree gloabal is PSMI/UI/MI/S MVPN instantiation of the tree how would that even work. There is some XR knob I am missing to get this working and is key for any mLDP MVPN profiles Verizon will be using. 6.4<https://urldefense.com/v3/__https:/tools.ietf.org/html/rfc7473*section-6.4__;Iw!!NEt6yMaO-gk!WnhQxBymsn_Oi0zcyNKYUSvz76hpDCbxFiywetIwJmmQDOeIEnDiWEniowQKaRs$>. Disabling Prefix-LSPs on an mLDP-only Session Assume that LSR1 and LSR2 have formed an LDP session to exchange mLDP state only. In typical deployments, LSR1 and LSR2 also exchange bindings for IP (unicast) prefixes upon mLDP session, which is unnecessary and wasteful for an mLDP-only LSR. Using the procedures defined earlier, an LSR can indicate its disinterest in Prefix-LSP application state to its peer upon session establishment time or dynamically later via an LDP capabilities update. In reference to Section 3.1<https://urldefense.com/v3/__https:/tools.ietf.org/html/rfc7473*section-3..1__;Iw!!NEt6yMaO-gk!WnhQxBymsn_Oi0zcyNKYUSvz76hpDCbxFiywetIwJmmQDOeIEnDiWEnivJQNlUY$>, the peer disables the advertisement of any state related to IP Prefix FECs, but it still advertises IP address bindings that are required for the correct operation of mLDP. -Rishabh On Sat, Mar 21, 2020 at 11:11 AM Gyan Mishra <hayabusa...@gmail.com<mailto:hayabusa...@gmail.com>> wrote: > > > Thank you Rishabh! > > I will contact you regarding Cisco specific questions. > > The questions apply to multicast support with SR-MPLS and are not vendor > specific however I am using Cisco as an example. > > From a IETF standards perspective, I believe the one question that this > thread is related is with multicast SR-MPLS support use case where you are > migrated to SR-MPLS and LDP has been removed from the SP core. > > In this customer use case where the customer does not want to use RSVP TE or > IR due to replication processing overhead in a distributed model what options > are available for multicast support. > > The SR replication segment tree sid draft states that it can only be used for > PCE centralized controller model. > > I am guessing BIER maybe an option? > > Any other options for operators? > > Kind regards > > Gyan > > On Sat, Mar 21, 2020 at 1:49 PM Rishabh Parekh > <risha...@gmail.com<mailto:risha...@gmail.com>> wrote: >> >> Gyan, >> These questions are implementation specific and should be addressed >> off the mailing list. Please contact me at >> ripar...@cisco.com<mailto:ripar...@cisco.com>. >> >> -Rishabh >> >> On Fri, Mar 20, 2020 at 6:41 PM Gyan Mishra >> <hayabusa...@gmail.com<mailto:hayabusa...@gmail.com>> wrote: >> > >> > >> > Daniel & Authors >> > >> > I had a question related to the draft related to lab POC testing. >> > >> > https://datatracker.ietf.org/doc/draft-voyer-spring-sr-replication-segment/<https://urldefense.com/v3/__https:/datatracker.ietf.org/doc/draft-voyer-spring-sr-replication-segment/__;!!NEt6yMaO-gk!WnhQxBymsn_Oi0zcyNKYUSvz76hpDCbxFiywetIwJmmQDOeIEnDiWEnihata2k0$> >> > >> > In the draft it states that SR replication using Tree SID to replication >> > to leafs on a tree is only supported with a centralized PCE controller >> > based model using BGP LS. >> > >> > I have an SR-MPLS Cisco VIRL POC test bed using XRV9000 nodes 7.0.1 using >> > ISIS SR extensions where I have L3 vpn overlay and everything is working >> > very well from a unicast perspective. No issues. >> > >> > I have LDP still enabled but via "SR-Prefer" am using SR-MPLS forwarding >> > plane. I kept LDP enabled so I can use mLDP for LMDT label switched trees >> > for multicast and technically that all MVPN procedures RFC 6513 6514 encap >> > tunnel types should work for p-tree using mLDP forwarding plane for >> > multicast while SR-MPLS is being used for unicast... >> > >> > I can get the LMDT core tree default and data mdt to build for MP2MP or >> > P2MP tree but cannot get on the FEC root the MRIB state to build.. Not >> > sure why? >> > >> > Any ideas. Is there anything special I have to do for multicast to use >> > the ldp mLDP extension data plane and not the SR-MPLS data plane. >> > >> > I think what's happening is at the data plane forwarding level SR-MPLS >> > data plane is being used instead of mLDP. >> > >> > I have a bunch of SR-TE policies in place with candidate dynamic and >> > static ERO paths and that works well coloring the VRF steering. >> > >> > I was wondering if I can use SR-TE binding Sid with Static ERO loose path >> > using prefix SID of egress PE to replicate to and build P2MP tree >> > instantiation via SR-TE. >> > >> > Is that possible? >> > >> > Kind regards >> > >> > Gyan >> > >> > >> > -- >> > >> > Gyan Mishra >> > >> > Network Engineering & Technology >> > >> > Verizon >> > >> > Silver Spring, MD 20904 >> > >> > Phone: 301 502-1347 >> > >> > Email: gyan.s..mis...@verizon.com<mailto:gyan.s..mis...@verizon.com> >> > >> > >> > >> > _______________________________________________ >> > spring mailing list >> > spring@ietf.org<mailto:spring@ietf.org> >> > https://www.ietf.org/mailman/listinfo/spring<https://urldefense.com/v3/__https:/www.ietf.org/mailman/listinfo/spring__;!!NEt6yMaO-gk!WnhQxBymsn_Oi0zcyNKYUSvz76hpDCbxFiywetIwJmmQDOeIEnDiWEniD2y-tsc$> > > -- > > Gyan Mishra > > Network Engineering & Technology > > Verizon > > Silver Spring, MD 20904 > > Phone: 301 502-1347 > > Email: gyan.s.mis...@verizon.com<mailto:gyan.s.mis...@verizon.com> > > > -- Gyan Mishra Network Engineering & Technology Verizon Silver Spring, MD 20904 Phone: 301 502-1347 Email: gyan.s.mis...@verizon.com<mailto:gyan.s.mis...@verizon.com> -- Gyan Mishra Network Engineering & Technology Verizon Silver Spring, MD 20904 Phone: 301 502-1347 Email: gyan.s.mis...@verizon.com<mailto:gyan.s.mis...@verizon.com>
_______________________________________________ spring mailing list spring@ietf.org https://www.ietf.org/mailman/listinfo/spring
