BESS WG Y. Wang Internet-Draft B. Song Intended status: Standards Track ZTE Corporation Expires: December 9, 2020 June 7, 2020 Context Label for MPLS EVPN draft-wang-bess-evpn-context-label-01 Abstract EVPN is designed to provide a better VPLS service than [RFC4761] and [RFC4762], and EVPN indeed introduced many new features which couldn't be achieved in those old VPLS implementions. But EVPN didn't inherit all features of old VPLS, and a few issues arises for EVPN only. Some of these issues can be imputed to the MP2P nature of EVPN labels. The PW label in old VPLS is a label for P2P VC, so it contains more context than a identifier in dataplane for it's VSI instance.But the EVPN label just identifies it's VSI instnace and it can't stand for the ingress PE in dataplane. So the following issues arises with MPLS EVPN service: MPLS EVPN statistics can't be done per ingress PE. MPLS EVPN can't support hub/spoke use case which the spoke PE can only connect to each other by the hub PE. MPLS EVPN can't support AR REPLICATOR. MPLS EVPN can't support anycast SR-MPLS tunnel on the SPE nodes. This document introduces a compound label stack to take advantage of both P2P VC and MP2P evpn labels. Status of This Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at https://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any Wang & Song Expires December 9, 2020 [Page 1] Internet-Draft EVPN Context Label June 2020 time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on December 9, 2020. Copyright Notice Copyright (c) 2020 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1. Terminology and Acronyms . . . . . . . . . . . . . . . . . . 3 2. Problem Statement . . . . . . . . . . . . . . . . . . . . . . 3 3. Context VC Infrastructure . . . . . . . . . . . . . . . . . . 4 3.1. The Shared Context VCs . . . . . . . . . . . . . . . . . 4 3.1.1. Signalling for Shared Context VCs . . . . . . . . . . 5 3.2. The per-EVI Context VCs . . . . . . . . . . . . . . . . . 6 3.2.1. Signalling for per-EVI Context VCs . . . . . . . . . 6 4. Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4.1. Solution for spoke PE isolating on hub PE . . . . . . . . 7 4.2. Solution for per ingress statistics . . . . . . . . . . . 8 4.2.1. Signalling using Context Label Space ID Extended Community . . . . . . . . . . . . . . . . . . . . . . 8 4.2.2. Signalling using Label of Context-Label-Space Extended Community . . . . . . . . . . . . . . . . . 9 4.3. Solution for AR REPLICATOR in MPLS EVPN . . . . . . . . . 9 4.4. Solution for anycast tunnel usage on SPE . . . . . . . . 10 5. Security Considerations . . . . . . . . . . . . . . . . . . . 11 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 11 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 11 8. Normative References . . . . . . . . . . . . . . . . . . . . 11 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 13 Wang & Song Expires December 9, 2020 [Page 2] Internet-Draft EVPN Context Label June 2020 1. Terminology and Acronyms This document uses the following acronyms and terms: BUM - Broadcast, Unknown unicast, and Multicast. CE - Customer Edge equipment. OPE - Originating PE - the original Router of an EVPN route. PE - Provider Edge equipment. ORIP - Originating Router's IP address. PTA - PMSI Tunnel Attribute. IR - Ingress Replication. AR - Assisted Replication. IR PTA - PMSI Tunnel Attribute with tunnel-type = IR. AR PTA - PMSI Tunnel Attribute with tunnel-type = AR. IRL - Ingress Replication List, the list for Ingress-Replication BUM packets forwarding. LS - Label Space. CLS - Context Label Space. 2. Problem Statement EVPN is designed to provide a better VPLS service than RFC4761/ RFC4762, and EVPN indeed introduced many new features which couldn't be achieved in those old VPLS implemention.But EVPN didn't inherit all features of old VPLS, and a few issues arises for EVPN only. Some of these issues can be imputed to the MP2P nature of EVPN labels. The PW label in old VPLS is a label for P2P VC, so it contains more context than a identifier in dataplane for it's VSI instance. But the EVPN label just identifies it's VSI instnace and it can't stand for the ingress PE in dataplane. So the following issues arises with MPLS EVPN service: MPLS EVPN statistics can't be done per ingress PE. All flows from remote PEs share the same statistics on egress PE, because they share Wang & Song Expires December 9, 2020 [Page 3] Internet-Draft EVPN Context Label June 2020 the same EVPN label and the egress PE can't pick them out in the dataplane. MPLS EVPN can't support hub/spoke usecase, where the spoke PEs can only connect to each other through the hub PE. Especially when at least two of the spoke PEs are connected to a common route reflector. MPLS EVPN can't work as an AR-REPLICATOR. Because the AR-REPLICATOR will apply replication for the ingress AR-LEAF too. But a packet shoud not be sent back to the AR-LEAF where it is received from. MPLS EVPN SPE cannot make use of SR-MPLS anycast tunnel because the two SPEs of the anycast tunnel will assign different EVPN labels for the same EVPN route. So this document introduces an compound label stack to take advantage of both P2P VC and MP2P evpn labels. 3. Context VC Infrastructure In order to add as much context as old VPLS to EVPN data packet, We can construct a infrastructure by a full-mesh of context VCs among the EVPN PEs. Take the context VCs between PE-i and PE-j as an example, VC-ij is the context VC from PE-i to PE-j, and VC-ji is the context VC from PE-j to PE-i. The VC-ij identifies the PE-i node on PE-j. The VC-ji identifies PE-j node on PE-i. The VC-label for VC-ij is called as L-ij, and the VC-label for VC-ji is called as L-ji. So the PE-i can push the L-ij in the EVPN data packet for PE-j to distinguish the packet of PE-i from other data packets. Because the L-ij identifies the ingress PE of the data packet. There are two styles of context VC in this draft. One style is named as shared context VC, the other style is named per-EVI context VC. 3.1. The Shared Context VCs The shared context VCs are dedicated to identify the context for a data packet while the EVPN label still identifies the EVPN instance. Note that typically a shared context VC can be shared by all the EVPN instances between it's ingress PE and egress PE. In other words, we don't have to construct a dedicated mesh of context VCs for each specified EVPN service. So we called the shared context VCs as a common infrastructure for those EVPN services. Wang & Song Expires December 9, 2020 [Page 4] Internet-Draft EVPN Context Label June 2020 3.1.1. Signalling for Shared Context VCs The VCs of a context VC infrastructure are set up by a context VC container, the container implements a VC signalling to set up the VCs. There are two existing signalling protocol can be reused to set up context VCs for a context VC container. 3.1.1.1. Kompella Signalling for context VC The signalling used by a Kompella VPLS instance per [RFC4761] can also be used by a context VC container. Different from the Kompella VPLS instance, a context VC container only use the signalling to set up the context VCs. They are the same in signalling but different in dataplane. Take the PW between PE-i and PE-j as an example, it is constructed by VC-ij and VC-ji, and none of the two context VCs will identify a MAC-VRF. In other words the PW is a context PW. Note that the context VC containers don't have a MAC-VRF or a MAC- table, they are just containers for context VC. 3.1.1.2. SR-MPLS signalling for context VC SR-MPLS signalling is very similar to the singleton pattern of Kompella VPLS, in spite of their different data plane and service procedure. The SID is similar to the VE-ID, the SRGB is similar to the label block. So the constructed LSPs of the SR-MPLS signalling can be reinterpreted as context VCs in another label space named S. These context VCs use the same label values as those SR-LSPs but they are constructed at the same time in different label spaces. Take the VC- ij as an example, its label value L-ij is the same as the SID label for PE-i in PE-j's SRGB. But the VC-ij are constructed in the context label space S which is identified by a static label. it is not constructed in the same label space with that SID label. The context VC signalling may be [RFC8665], [RFC8666], [RFC8667]. The context VC may be established along with SR-LSPs. Wang & Song Expires December 9, 2020 [Page 5] Internet-Draft EVPN Context Label June 2020 +---------------------------------+ | underlay ethernet header | +---------------------------------+ | PSN tunnel label | +---------------------------------+ | EVPN label | +---------------------------------+ | Static Label for Label Space S | +---------------------------------+ | Context VC Label | +---------------------------------+ | overlay ethernet or IP header | +---------------------------------+ Figure 1: Encapsulation of Context VC Label in Context Label Space Note that the static label S is the context label for L-ij, while the L-ij is the context label for the data packet. 3.2. The per-EVI Context VCs The per-EVI context VCs are used to identify both the context (typically the ingress-PE) and the EVPN instance for a data packet at the same time. In other words, we have to construct a dedicated set of per-EVI context VCs for each specified EVPN service. 3.2.1. Signalling for per-EVI Context VCs The IMET route per [RFC7432] have a corresponding route-type in MVPN. It is, in effect, the Intra-AS I-PMSI route per [RFC6514]. But an IMET route with Ingress Replication (IR) tunnel type PMSI Tunnel Attribute (PTA) doesn't need a responding Leaf A-D route. The Leaf A-D route per [I-D.ietf-bess-evpn-bum-procedure-updates] is required for P2MP PTA only. In this draft, we use the Leaf A-D route with IR- PTA to construct per-EVI context-VCs. 3.2.1.1. Construct Leaf A-D Route for IR PE1 will construct a Leaf A-D route with IR-PTA for EVI1 in response to an IMET route R1 with IR-PTA. The IMET route R1 is received from PE2 previously. The key fields of the IMET route is included in the "Route Type specific" fields of the Leaf A-D route (say R2) along with the ORIP of PE1 itself. We call the ORIP of PE1 itself as the Leaf A-D route's "self-ORIP" in order to distinguish it from the "Route Type specific" ORIP. So the key fields of the Leaf A-D route is per basis. Wang & Song Expires December 9, 2020 [Page 6] Internet-Draft EVPN Context Label June 2020 The MPLS label field in the IR-PTA of the Leaf A-D route is allocated per basis in per-platform label space. So the per-EVI context VC can identify the EVI1 too. Note that PE1 may already advertise an IMET route R3 to PE2 before the advertisement of above Leaf A-D route. Note that the MPLS label field in the IR-PTA of R2 (Leaf A-D) may be the same label in the IR- PTA of R3 (IMET) either. In such case, the IR-PTA is included in the Leaf A-D route along with a "Context Label Space (CLS) ID Extended Community" per [I-D.ietf-bess-mvpn-evpn-aggregation-label]. The ID- type field of the CLS-ID EC is 0, the ID-Value field of the CLS-ID EC is a label of "Shared Context VC" Label. 3.2.1.2. Construct Ingress Replication List by Leaf A-D Route PE2 receives the responding Leaf A-D route (say R2) of the IMET route R1 which is previously advertised by itself, and PE2 preiously received an IMET route R3 with the same ORIP as the self-ORIP of R2 . Given that R1,R2 and R3 both have a IR-PTA, PE2 SHOULD use R2 to construct the Ingress Replication List (IRL) item for PE1 instead, and R3 will not used to construct the IRL-item for PE1 from then on. Note that when R2 included a CLS-ID EC, the ID-value of the CLS-ID EC will be used as outgoing label by the IRL-item. It will be used as the context label of the MPLS label of the IR-PTA. 4. Solutions 4.1. Solution for spoke PE isolating on hub PE PEs1--------RR1--------PEh---------RR2--------PEs3 / PEs2-------/ Figure 2: Hub PE and Spoke PEs Now take above use case for example, there are three spoke PEs and one hub PE. The spoke PEs are PEs1, PEs2 and PEs3. The hub PE is PEh. Two of the spoke PEs (PEs1 and PEs2) are connected to the same RR group and the third one connects to another RR group. Although we can advertise different EVPN labels for different RR groups, we can't advertise different EVPN labels for PEs1 and PEs2. But PEh can request PEs1 or PEs2 to push the label of the context VC from it to PEh. Benefit from the context VC label, PEh can Wang & Song Expires December 9, 2020 [Page 7] Internet-Draft EVPN Context Label June 2020 distinguish where the packet from, in other words, PEh can decide where the packet can't be sent to. The signaling for the hub PE to request the spoke PE to push the context VC label will be added in future versions. Note that although PEs1 and PEs2 can receive EVPN routes from each other they won't import these routes because of the hub/spoke behaviors. 4.2. Solution for per ingress statistics The EVPN label is allocated from per-platform label space, and it identifies the EVPN instance as per [RFC7432]. But it also identifies a context label space LS1. 4.2.1. Signalling using Context Label Space ID Extended Community The signalling in Section 3.2.1.1 with CLS-ID EC will be used. But the ID-value in CLS-ID EC is the EVPN Label, and the IR-PTA label of the Leaf A-D route will be allocated in LS1 per TPE basis, and it is actually the context VC label. So the context VC label need to be pushed to the label stack before the EVPN Label. Such encapsulation is illustrated as the following figure: +---------------------------------+ | underlay ethernet header | +---------------------------------+ | PSN tunnel label | +---------------------------------+ | EVPN label | +---------------------------------+ | Context VC Label | +---------------------------------+ | overlay ethernet or IP header | +---------------------------------+ Figure 3: Encapsulation of Context VC Label for EVPN Payload Note that the Context VC Label here is not the CLS-ID of the EVPN Label. But the EVPN label is the CLS-ID of the Context VC Label. And the label space LS1 may be actually the per-platform label space. Wang & Song Expires December 9, 2020 [Page 8] Internet-Draft EVPN Context Label June 2020 4.2.2. Signalling using Label of Context-Label-Space Extended Community Note that when the label space LS1 is actually the per-platform label space, and PE1 send a Leaf A-D route with CLS-ID EC to PE2, but PE2 don't recognize the CLS-ID EC, then PE2 will encapsulate the context VC label without the EVPN label. This will cause packet drop. So we introduce a new EC called Label of CLS (LoCLS) EC, the LoCLS EC has the same format as the CLS-ID EC except for a different code- point of it's "sub-type" field. The ID-Value of the LoCLS EC is a MPLS label in a context label space identified by the PTA label. And the MPLS label in LoCLS EC will be pushed to the label stack before the PTA label by the ingress PE. Typically, the MPLS label of the LoCLS EC is a downstream assigned label, which means that it will be used as outgoing label by the PE receiving the LoCLS EC, not as incomming label. When constructing the Leaf A-D route, the IR-PTA label is the EVPN Label, as per [RFC7432]. But the ID-value in LoCLS ES is a label which is allocated in LS1 per TPE basis, and it is actually the context VC label. So the context VC label need to be pushed to the label stack before EVPN Label (which identifies LS1) on ingress PEs. Note that when PE1 send a Leaf A-D route with LoCLS EC to PE2, but PE2 don't recognize the LoCLS EC, then PE2 will encapsulate the EVPN label without the inner context label. This will work as well as [RFC7432], although the per-ingress statistics can't be applied. Note that the LoCLS ECs (for different EVIs) received from the same TPE will be the same label, So we can select a single EVI to use the Leaf A-D route with LoCLS EC. This EVI is called as administrating EVI (admin-EVI). The context VC label carried in the Leaf A-D routes of the admin-EVI will be used for the IMET routes with the same ORIP in all other ordinary EVIs in such case. Note that all other ordinary EVIs don't use the Leaf A-D routes with IR-PTA, they use ordinary IMET routes instead. The admin-EVI need to be configured on all EVPN-PEs in such case. 4.3. Solution for AR REPLICATOR in MPLS EVPN LEAF1--------REPLICATOR1--------RNVE1 / LEAF2-----------/ Figure 4: AR REPLICATOR in MPLS EVPN Wang & Song Expires December 9, 2020 [Page 9] Internet-Draft EVPN Context Label June 2020 When REPLICATOR1 node recieves an IMET Route with AR-role = AR-LEAF from LEAF1 node, REPLICATOR1 SHOLD respond to it with an Leaf A-D route with AR-PTA. The MPLS label field of the AR-PTA (say AR-PTA Label) will be allocated following the same rules as the IR-PTA Label in Section 3.2.1.1. When ALEAF1 receives above Leaf A-D route, the Leaf A-D route is treated as a Replicator-AR route for the same ORIP, and then the control-plane procedures works following [I-D.ietf-bess-evpn-optimized-ir]. When REPLICATOR1 receives data packets from the AR-PTA Label, REPLICATOR1 will do source-squelching for LEAF1 which means that these data packets will not be forwarded back to LEAF1. Note that the old Replicator-AR route which is in terms of IMET route will not be used by MPLS EVPN AR-REPLICATOR. Because that the Leaf A-D routes will take it's place per AR-LEAF basis. But the old Regular-IR route can still be used by MPLS EVPN AR-REPLICATORs. 4.4. Solution for anycast tunnel usage on SPE /--------SPE1-------\ TPE1 TPE2 \--------SPE2-------/ Figure 5: SPE with Anycast Tunnel Now take above use case for example, the two SPEs are the egress nodes of an anycast SR-MPLS tunnel. The anycast SR-MPLS tunnel is used to transport flows from TPE1 to either SPE1 or SPE2 according to load balancing procedures. So SPE1 and SPE2 have to advertise the same EVPN label independently for a given EVPN route. In fact, SPE1 and SPE2 can simply inherit the EVPN label from TPE2, and they advertise it to TPE1 along with a context VC label. The context VC label is for the context VC from TPE2 to SPE1 or SPE2. We can make the VC labels from TPE2 to SPE1 and SPE2 have the same value through configuring. And the label stack on the anycast SR-MPLS tunnel is constructed as the following: Wang & Song Expires December 9, 2020 [Page 10] Internet-Draft EVPN Context Label June 2020 +---------------------------------+ | underlay ethernet header | +---------------------------------+ | Anycast SR-MPLS tunnel label | +---------------------------------+ | Static Label for Label Space S | +---------------------------------+ | Context VC Label | +---------------------------------+ | EVPN label | +---------------------------------+ | overlay ethernet or IP header | +---------------------------------+ Figure 6: Encapsulation of Context VC Label for EVPN Label Note that the context VC is also constructed in a context label space, the label space is identified by a static label. And the context label space is identified by the same label on all PEs of the service domain. so the label stacks on the anycast tunnel are the same for SPE1 and SPE2. SPE1/SPE2 will perform ILM lookup for the EVPN label in the label space identified by the context VC label. 5. Security Considerations This section will be added in future versions. 6. IANA Considerations The IANA considerations for LoCLS EC in Section 4.2.2 will be added in future versions. 7. Acknowledgements The authors would like to thank the following for their comments and review of this document: Benchong Xu. 8. Normative References Wang & Song Expires December 9, 2020 [Page 11] Internet-Draft EVPN Context Label June 2020 [I-D.ietf-bess-evpn-bum-procedure-updates] Zhang, Z., Lin, W., Rabadan, J., Patel, K., and A. Sajassi, "Updates on EVPN BUM Procedures", draft-ietf- bess-evpn-bum-procedure-updates-08 (work in progress), November 2019. [I-D.ietf-bess-evpn-optimized-ir] Rabadan, J., Sathappan, S., Lin, W., Katiyar, M., and A. Sajassi, "Optimized Ingress Replication solution for EVPN", draft-ietf-bess-evpn-optimized-ir-06 (work in progress), October 2018. [I-D.ietf-bess-mvpn-evpn-aggregation-label] Zhang, Z., Rosen, E., Lin, W., Li, Z., and I. Wijnands, "MVPN/EVPN Tunnel Aggregation with Common Labels", draft- ietf-bess-mvpn-evpn-aggregation-label-03 (work in progress), October 2019. [RFC4761] Kompella, K., Ed. and Y. Rekhter, Ed., "Virtual Private LAN Service (VPLS) Using BGP for Auto-Discovery and Signaling", RFC 4761, DOI 10.17487/RFC4761, January 2007, . [RFC6514] Aggarwal, R., Rosen, E., Morin, T., and Y. Rekhter, "BGP Encodings and Procedures for Multicast in MPLS/BGP IP VPNs", RFC 6514, DOI 10.17487/RFC6514, February 2012, . [RFC7432] Sajassi, A., Ed., Aggarwal, R., Bitar, N., Isaac, A., Uttaro, J., Drake, J., and W. Henderickx, "BGP MPLS-Based Ethernet VPN", RFC 7432, DOI 10.17487/RFC7432, February 2015, . [RFC8665] Psenak, P., Ed., Previdi, S., Ed., Filsfils, C., Gredler, H., Shakir, R., Henderickx, W., and J. Tantsura, "OSPF Extensions for Segment Routing", RFC 8665, DOI 10.17487/RFC8665, December 2019, . [RFC8666] Psenak, P., Ed. and S. Previdi, Ed., "OSPFv3 Extensions for Segment Routing", RFC 8666, DOI 10.17487/RFC8666, December 2019, . [RFC8667] Previdi, S., Ed., Ginsberg, L., Ed., Filsfils, C., Bashandy, A., Gredler, H., and B. Decraene, "IS-IS Extensions for Segment Routing", RFC 8667, DOI 10.17487/RFC8667, December 2019, . Wang & Song Expires December 9, 2020 [Page 12] Internet-Draft EVPN Context Label June 2020 Authors' Addresses Yubao Wang ZTE Corporation No. 50 Software Ave, Yuhuatai Distinct Nanjing China Email: yubao.wang2008@hotmail.com Bing Song ZTE Corporation No. 50 Software Ave, Yuhuatai Distinct Nanjing China Email: song.bing@zte.com.cn Wang & Song Expires December 9, 2020 [Page 13]