Skip to main content

Reduction of EVPN C-MAC Overload
draft-wang-bess-evpn-cmac-overload-reduction-01

The information below is for an old version of the document.
Document Type
This is an older version of an Internet-Draft whose latest revision state is "Expired".
Author Yubao(Bob) Wang
Last updated 2020-06-30 (Latest revision 2020-05-17)
RFC stream (None)
Formats
Additional resources
Stream Stream state (No stream defined)
Consensus boilerplate Unknown
RFC Editor Note (None)
IESG IESG state I-D Exists
Telechat date (None)
Responsible AD (None)
Send notices to (None)
draft-wang-bess-evpn-cmac-overload-reduction-01
BESS WG                                                          Y. Wang
Internet-Draft                                           ZTE Corporation
Intended status: Standards Track                             1 July 2020
Expires: 2 January 2021

                    Reduction of EVPN C-MAC Overload
            draft-wang-bess-evpn-cmac-overload-reduction-01

Abstract

   When there are too many customer-MACs (C-MACs), the RRs and/or ASBRs
   will be overloaded by the RT-2 routes for these MACs according to
   [RFC7432].  This issue can be simply solved by making the remote
   C-MAC entries learnt via data-plane MAC learning (like what PBB VPLS
   have done since [RFC7041]) rather than received from RT-2 routes.
   This simplified solution will works as well as PBB VPLS.  But this
   simplified solution will lose many important features that based on
   the ESI concept.  Because the ingress-ESI can't be learnt via data-
   plane MAC learning at the egress PE.  So when the data packets is
   forwarded following these MAC entries, they can't benefit from the
   EAD/EVI routes as per RFC7432.  So the All-Active Redundancy mode for
   ES can't be supported.  This make the simplified solution can't work
   as well as PBB EVPN ([RFC7623]).

   This document proposes some new extensions to [RFC7432] to achieve
   all-active mode ES redundancy on TPEs and reduce the C-MAC loads for
   RRs and ASBRs at the same time.  The new solution will work even more
   better than PBB EVPN under the help of these extensions, especially
   when there is no deployment of MPLS dataplane.

   In Section 6.2, this draft is compared with PBB EVPN and other
   solutions . These solutions are PBB EVPN, PBB VPLS, VXLAN([RFC7348]).

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
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

Wang                     Expires 2 January 2021                 [Page 1]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   This Internet-Draft will expire on 2 January 2021.

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.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   3
     1.1.  Terminology . . . . . . . . . . . . . . . . . . . . . . .   4
   2.  Requirements  . . . . . . . . . . . . . . . . . . . . . . . .   5
     2.1.  No C-MAC Awareness in the Backbone  . . . . . . . . . . .   6
     2.2.  EVPN IRB Support  . . . . . . . . . . . . . . . . . . . .   6
     2.3.  Unified Encapsulation per Scenario  . . . . . . . . . . .   6
     2.4.  ESI Features Remain Supported . . . . . . . . . . . . . .   6
     2.5.  Flexible Multi-homing Remains Supported . . . . . . . . .   7
     2.6.  C-MAC Address Learning and Confinement  . . . . . . . . .   7
     2.7.  No C-MAC Flushing for All-Active ESes . . . . . . . . . .   7
     2.8.  Independent C-MAC Flushing for Single-Active ESes . . . .   7
     2.9.  Independent Convergency per <ESI, EVI>  . . . . . . . . .   7
     2.10. Route Aggregation and Default Route in Backbone . . . . .   8
   3.  Solution Overview . . . . . . . . . . . . . . . . . . . . . .   8
     3.1.  Common Overview . . . . . . . . . . . . . . . . . . . . .   8
     3.2.  Pseudo PBB EVPN . . . . . . . . . . . . . . . . . . . . .   8
     3.3.  VXLAN Solution Overview . . . . . . . . . . . . . . . . .  10
     3.4.  MPLS Solution Overview  . . . . . . . . . . . . . . . . .  12
     3.5.  SRv6 Solution Overview  . . . . . . . . . . . . . . . . .  13
     3.6.  Comparisons of Relative Concepts  . . . . . . . . . . . .  14
   4.  Dataplane-specific Procedures . . . . . . . . . . . . . . . .  15
     4.1.  Packet Walkthrough  . . . . . . . . . . . . . . . . . . .  15
     4.2.  Pseudo PBB-EVPN . . . . . . . . . . . . . . . . . . . . .  16
     4.3.  VXLAN over IP-VRF . . . . . . . . . . . . . . . . . . . .  17
     4.4.  NVO3-specific EVPN-lite Procedures  . . . . . . . . . . .  17
     4.5.  MPLS-specific EVPN-lite Procedures  . . . . . . . . . . .  18
     4.6.  SRv6-specific EVPN-lite Procedures  . . . . . . . . . . .  19
   5.  Other Considerations  . . . . . . . . . . . . . . . . . . . .  20
     5.1.  ESI Indicator Advertisement Optimization  . . . . . . . .  20
     5.2.  C-MAC Flush Notification Procedure  . . . . . . . . . . .  20

Wang                     Expires 2 January 2021                 [Page 2]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

     5.3.  E-Tree Support Considerations . . . . . . . . . . . . . .  20
     5.4.  EVPN IRB Support Considerations . . . . . . . . . . . . .  21
     5.5.  Use End.ESI SID in MAC/IP Advertisement Routes  . . . . .  21
     5.6.  Hierarchical VPLS in EVPN-lite  . . . . . . . . . . . . .  21
   6.  Comparison with Other Solutions . . . . . . . . . . . . . . .  22
     6.1.  Questions . . . . . . . . . . . . . . . . . . . . . . . .  22
     6.2.  Summary Comparisons . . . . . . . . . . . . . . . . . . .  23
     6.3.  Detailed Comparisons with PBB EVPN  . . . . . . . . . . .  24
     6.4.  Detailed Comparisons with PBB VPLS  . . . . . . . . . . .  24
     6.5.  Detailed Comparisons with EVPN VXLAN  . . . . . . . . . .  24
   7.  Security Considerations . . . . . . . . . . . . . . . . . . .  24
   8.  IANA Considerations . . . . . . . . . . . . . . . . . . . . .  24
     8.1.  END.ESI SID . . . . . . . . . . . . . . . . . . . . . . .  24
     8.2.  Global Unique ESI-label in EAD per ES Route . . . . . . .  25
   9.  Acknowledgements  . . . . . . . . . . . . . . . . . . . . . .  25
   10. References  . . . . . . . . . . . . . . . . . . . . . . . . .  25
     10.1.  Normative References . . . . . . . . . . . . . . . . . .  25
     10.2.  Informative References . . . . . . . . . . . . . . . . .  26
   Author's Address  . . . . . . . . . . . . . . . . . . . . . . . .  26

1.  Introduction

   In [RFC7432], the C-MACs is advertised via RT-2 route.  This behavior
   is inheritted by [RFC8365] and [I-D.dawra-bess-srv6-services].  but
   in order to solve the C-MAC overload problem for RRs and ASBRs, we
   have to return to a PBB-like dataplane C-MAC learning procedures.

   We discuss all the requirements for a light-weight EVPN solution
   which pushes no C-MAC entries into the backbone network in Section 2.
   Note that some of these requirements is not supported well by PBB
   EVPN.

   In this document, the light-weight EVPN solutions are also called as
   EVPN-lite for short.  A total of five EVPN-lite solutions are
   proposed in Section 3.  These solutions are Pseudo PBB-EVPN, VXLAN
   over IP-VRF, VXLAN-based EVPN-lite, MPLS-based EVPN-lite, SRv6-based
   EVPN-lite.

   Note that the Pseudo PBB-EVPN solution is not a pragmatic solution,
   it is just a techniquely practicable solution for us to easily
   understand the equivalency between PBB EVPN and VXLAN over IP-VRF.
   Because the Pseudo PBB-EVPN is exactly the same as PBB EVPN in the
   control-plane, at the same time, the Pseudo PBB-EVPN is exactly the
   same as VXLAN over IP-VRF in the data-plane.

Wang                     Expires 2 January 2021                 [Page 3]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   Note that the VXLAN-based EVPN-lite and MPLS-based EVPN-lite are both
   a evolution of VXLAN over IP-VRF.  The former takes the MPLS factors
   off the VXLAN over IP-VRF solution, and becomes a pure VXLAN
   solution.  The latter takes the VXLAN factors off the VXLAN over IP-
   VRF solution, and becomes a pure MPLS solution.

   In order to compare these five solutions with [RFC7348] and [RFC7623]
   whose C-MAC entries are also not pushed into the backbone network,
   two terms are introduced in this document, because the comparisons
   need to be done in unified terminology.  One term is "Global ESI
   Indicator (GEI)", which is called as B-MAC in PBB EVPN.  The other
   term is "EVI's Global Dicreminator (EGD)", which is called as I-SID
   in PBB EVPN.

   Note that the EVI here corresponds to the I-Component of [RFC7623],
   not the B-Component.  In fact, there will be no B-components in some
   of the above seven solutions.

   Note that the GEI and EGD in different EVPN-lite solutions are very
   different.  The details will be described in Section 3.

   On the basis of GEI concept, then we define two route-types for EVPN-
   lite: The first route type is GEI/ES route, which is called as RT-2
   route in PBB EVPN.  The second route type is GEI/EVI route, which is
   called as EAD/EVI roue in [RFC7432].

   The details of these terms are described in Section 1.1.

1.1.  Terminology

   Most of the terminology used in this documents comes from [RFC7432]
   and [I-D.dawra-bess-srv6-services] except for the following:

   *  C-MAC: Customer MAC, it is the same as the C-MAC of PBB EVPN.

   *  ISID: a broadcast domain identifier in PBB I-Component.

   *  LDV: Local Discreminating Value.  It is similar to the Local
      Discreminating Value of type 3 ESI.

   *  GDV: Global Discreminating Value.  An identifier with global
      uniqueness.

   *  EGD: EVI-GDV, an EVI's Global Discreminator, it is a GDV for an
      EVI instance.  A EGD is used to idenfify an EVPN Instance (EVI) in
      data plane.  The EGD is a Global Discreminating Value (GDV) of
      that EVI, so it is also the abbreviation of EVI-GDV.  e.g.  The
      EGD of [RFC7348] is a global VNI.

Wang                     Expires 2 January 2021                 [Page 4]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   *  ESI Indicator: A Global ID for an ESI.  Note that different PE may
      assign different ESI-indicator for the same ESI, espacially when
      the ES redundancy mode is single-active.  e.g.  The ESI indicator
      of [RFC7623] is B-MAC.

   *  GEI: Global ESI Indicator.  It is the same as the "ESI Indicator"
      except for the emphasization to its global uniqueness.  A GEI is
      used in data plane to identify an ESI, because it have global
      uniqueness across the service domain of a corresponding EVPN
      Instance (EVI).  But an ESI may have a few GEIs, each for a TPE,
      espacially in the single-active mode of ES redundancy.  And in
      E-Tree scenarios, an ESI may have two GEIs on the same PE, one for
      Root ACs, one for Leaf ACs.  e.g.  The GEIs for an ESI of
      [RFC8317] is two B-MACs, one for root ACs, one for Leaf ACs.

   *  GEI/ES: The EVPN route which is used to advertise the relation
      between ESE and its GEI.  Note that the GEI/ES route is advertised
      per ESI basis on a specified PE.  In PBB EVPN, the GEI/ES route is
      the MAC Advertisement Route.  Note that different solutions may
      have different GEI/ES routes.

   *  EAD/EVI: An Ethernet A-D route per EVI.

   *  GEI/EVI: The EVPN route which is used to advertise the relation
      between <ESI/GEI, EVI> and its EVPN label and MPLS nexthops.  Note
      that in PBB EVPN, such route is not used.  Note that different
      solutions may have different GEI/EVI routes.

   *  ARG.EGD: The argument part of a SID of the End.ESI function is
      called as ARG.EGD, because the value of that argument will be a
      EGD.

   *  RT-2: MAC/IP Advertise Route.

   *  MAC Entry: An entry in the EVPN MAC table in data-plane.

   *  ESI IP: An End.ESI SID with its Argument part being set to zero.

   *  VXLAN EVPN: EVPN per [RFC8365].

   *  EVPN VXLAN: A broadcast domain per [RFC7348], but use IMET routes
      of [RFC8365] to construct VXLAN tunnels.

2.  Requirements

   EVPN C-MAC Reduction should be provided together with the following
   requirements:

Wang                     Expires 2 January 2021                 [Page 5]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

2.1.  No C-MAC Awareness in the Backbone

   In typical operation, an EVPN PE sends a BGP MAC Advertisement route
   per C-MAC address.  In certain applications, this poses scalability
   challenges, as is the case in data center interconnect (DCI)
   scenarios where the number of virtual machines (VMs), and hence the
   number of C-MAC addresses, can be in the millions.  This is called as
   C-MAC overload of DC Backbone.  In such scenarios, it is required to
   reduce the number of BGP MAC Advertisement routes by relying on a
   'EVPN-lite' scheme, as is provided by ESI and its equivalents (e.g.
   Pseudo B-MAC, ESI IP).

2.2.  EVPN IRB Support

   The PBB-VPLS/PBB-EVPN is not friendly to IRB usecase because of its
   complicated Protocol Stack, so it is used just in pure L2VPN usecase
   up to now in the industry.

   The solution should provider efficient forwarding performance in EVPN
   IRB use cases.

2.3.  Unified Encapsulation per Scenario

   PBB EVPN, especially the MPLS encapsulation of its B-VPLS, is
   typically not used in DC Scenario.  So we bring PBB and MPLS
   encapsulation to DC Backbone just due to the C-MAC overload problem.
   EVPN IRB is widely deplyed in DC scenarios, but PBB EVPN is not
   friendly for EVPN IRB use cases.  So we have to use different
   solutions in EVPN IRB and C-MAC reduction use cases.  We believe that
   if we choose VXLAN/Geneve data-plane, we will prefer to use the same
   data-plane in all use cases, e.g.  EVPN IRB, C-MAC reduction.  So it
   is necessary to make NVO3/MPLS/SRv6 EVPN to support Section 2.1 in
   order to provider a unified solution for data center and other
   secenarios.

2.4.  ESI Features Remain Supported

   Two redundancy modes are defined in [RFC7432].  They are All-Active
   mode and Single-Active mode.

   In All-active mode, the C-MAC movement among the different adjacent
   PE nodes of the same ESI should not be considered as C-MAC mobility.
   In Single-Active mode, such movements can be considered as C-MAC
   mobility.

Wang                     Expires 2 January 2021                 [Page 6]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

2.5.  Flexible Multi-homing Remains Supported

   Flexible multi-homing means that different ES instances can have
   different adjacent-PEs.  We call all the adjacent-PEs of the same ES
   instances as that ES's location-set in this document.  Flexible
   multi-homing means that different ES can have different location-set.

   For example, ES1's location-set is {PE1}, ES2's location-set is {PE2,
   PE3}, ES3's location-set is {PE1, PE3}, and ES4's location-set is
   {PE2,PE4}.

2.6.  C-MAC Address Learning and Confinement

   In EVPN, all the PE nodes participating in the same EVPN instance are
   exposed to all the C-MAC addresses learned by any one of these PE
   nodes because a C-MAC learned by one of the PE nodes is advertised in
   BGP to other PE nodes in that EVPN instance.  This is the case even
   if some of the PE nodes for that EVPN instance are not involved in
   forwarding traffic to, or from, these C-MAC addresses.  Even if an
   implementation does not install hardware forwarding entries for C-MAC
   addresses that are not part of active traffic flows on that PE, the
   device memory is still consumed by keeping record of the C-MAC
   addresses in the routing information base (RIB) table.  In network
   applications with millions of C-MAC addresses, this introduces a non-
   trivial waste of PE resources.  As such, it is required to confine
   the scope of visibility of C-MAC addresses to only those PE nodes
   that are actively involved in forwarding traffic to, or from, these
   addresses.

2.7.  No C-MAC Flushing for All-Active ESes

   Just as in [RFC7432], it is required to avoid C-MAC address flushing
   upon link, port, or node failure for All-Active multihomed segments.

2.8.  Independent C-MAC Flushing for Single-Active ESes

   Just as in [RFC7432], upon singel-active ES1's link or port failure,
   the C-MACs of other single-active ESes from the same PE will not be
   flushed.

2.9.  Independent Convergency per <ESI, EVI>

   When the physical port of an All-Active ES works well, but a single
   Ethernet Tag ID (ETI) of that ES fails, The traffic to that ETI of
   that ES will be re-routed to other adjacent PE of the same ES, but
   the traffic to other ETIs of the same ES will not be affected.

Wang                     Expires 2 January 2021                 [Page 7]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

2.10.  Route Aggregation and Default Route in Backbone

   The routes per ESIs can be aggregated in Backbone network.  Even the
   default route should be supported.

3.  Solution Overview

3.1.  Common Overview

   We assign a Global Discreminator EGD1 to an EVI instance EVI1, the
   EGD1 is a number consists of N bits.  We assign an ESI-indicator GEI1
   to ESI1 on PE1, and we assign an ESI-indicator GEI2 to ESI1 on PE2.
   We call the relationship between ESI1 and its two ESI-indicators as
   ESI1_GEI1 and ESI1_GEI2 respectively.  The EGD and GEIs MUST have
   global uniqueness in EVI1's service domain.

                                    +----------+
                      PE1           |          |
                 +-------------+    |          |
                 | ESI1_GEI1   |    |          |         PE3
                /|             |----|          |   +-------------+
               / |             |    | IP/MPLS  |   |             |
          LAG /  +-------------+    | Backbone |   |   ESI2_GEI3 |---CE2
      CE1=====                      |   with   |   |             |
              \  +-------------+    |   EVPN   |---|             |
               \ |             |    |   RRs    |   +-------------+
                \|             |----|   and    |
                 | ESI1_GEI2   |    |   SPEs   |
                 +-------------+    |          |
                      PE2           |          |
                                    +----------+

                    Figure 1: EVPN MAC Reduction Usecase

   We use IMET routes to build a broadcast-list.  The broadcast-list is
   used to forward BUM traffics.  The data-plane MAC learning for BUM
   traffics produces the first batch of C-MAC entries.  The subsequent
   C-MAC entries can be learnt from Unicast traffics and/or BUM
   traffics.  It is clear that we don't use MAC/IP routes to advertise
   C-MAC entries as usual, that is for fear that the RRs and/or ASBRs
   are overloaded by these C-MACs.

3.2.  Pseudo PBB EVPN

   Pseudo PBB EVPN means that we use the management and control plane of
   PBB EVPN but use the data plane of EVPN VXLAN.

Wang                     Expires 2 January 2021                 [Page 8]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   The BMACs of Pseudo PBB EVPN is called as Pseudo BMACs in this draft,
   the format of a pseudo-BMAC is illustrated as the following:

       |            24 bits          |         24 bits             |
       +----+----+----+----+----+----+----+----+----+----+----+----+
       |         OUI = PBB2IP        |  NIC-ID = ESI IP's Host ID  |
       +----+----+----+----+----+----+----+----+----+----+----+----+

                  Figure 2: GEI Format of Pseudo PBB-EVPN

   The OUI part of a pseudo-BMAC is a constant value called PBB2IP.  We
   suppose that PBB2IP is 0x20000A, so that the lower four bytes of the
   pseudo-BMAC is exactly right a IP address of the IPv4 prefix
   10.0.0.0/8.  When the Pseudo-BMAC is for an ESI x, that IP address is
   called as ESI x's ESI IP in Pseudo PBB-EVPN.  When the Pseudo-BMAC is
   for a PE a, that IP address is called as PE a's PE IP in Pseudo PBB-
   EVPN.  The Pseudo-BMAC is not used in the dataplane, that's why it is
   called as pseudo-BMAC.

   Note that when the NIC-specific part of a pseudo-BMAC has the same
   value with the host ID part of an IPv4 address, We say that this IPv4
   address equals that pseudo-BMAC, or this IPv4 address in data plane
   is in the name of that pseudo-BMAC, or just "IP=BMAC".

   The ESI/PE IP will be used in the dataplane in the name of their
   Pseudo-BMAC.  So it won't be so surprising that we don't use PBB
   encapsulation in dataplane.  In fact we use VXLAN over IP-VRF data
   plane in Pseudo PBB-EVPN, the encapsulation is illustrated as the
   following:

                +----------------------------------------+
                | IP/MPLS Header (Backbone IP-VRF label) |
                +----------------------------------------+
                | IPv4 Header (SIP=S-BMAC, DIP=D-BMAC)   |
                +----------------------------------------+
                | UDP Header                             |
                +----------------------------------------+
                | VXLAN Header (VNI=EGD=I-SID)           |
                +----------------------------------------+
                | Ethernet Header                        |
                +----------------------------------------+
                | Ethernet Payload                       |
                +----------------------------------------+
                | Ethernet FCS                           |
                +----------------------------------------+

                  Figure 3: Pseudo PBB-EVPN Encapsulation

Wang                     Expires 2 January 2021                 [Page 9]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   It is clearly illustrated that the VXLAN header is in the name of old
   PBB Header, the VNI is in the name of old I-SID, the underlay source/
   destination IP of VXLAN header is in the name of source/destination
   BMAC, the backbone IP-VRF is in the name of old B-VPLS instance.

   As a sequence of the above settings, the B-MAC based forwarding in
   old B-VPLS is replaced with IP forwarding in IP-VRF instances.

   Note that the encapsulation in Figure 3 is completely the same as
   [RFC7348] except for the MPLS encapsulation.

   Note that the GEI in the control-plane is B-MAC, the GEI in the data-
   plane is ESI-IP.  And the EGD in control-plaen is I-SID, the EGD in
   data-plane is VNI.

   The GEI/ES route is the RT-2 route of PBB-EVPN.  There is no GEI/EVI
   route in Pseudo PBB-EVPN, just as there is no GEI/EVI route in PBB
   EVPN.

3.3.  VXLAN Solution Overview

   Although the VXLAN encapsulation is used in Pseudo PBB-EVPN, the MPLS
   data plane is also needed, because that the B-Component of Pseudo
   PBB-EVPN is an IP-VRF.  So we introduce a completely VXLAN
   encapsulation in this section, it is called as VXLAN solution.

   In VXLAN solution, the GEI is composed of a VTEP IP and an ESI label.
   It is illustrated as the following:

       |           32 bits                     |      16 bits      |
       +----+----+----+----+----+----+----+----+----+----+----+----+
       |           VTEP IP                     |     ESI Label     |
       +----+----+----+----+----+----+----+----+----+----+----+----+

                         Figure 4: VXLAN GEI Format

   The VTEP IP is PE1/PE2/PE3's IP address, the ESI label is a local
   discreminating value (LDV) that is used to identify a ESI.  So the
   GEI has global uniqueness in the EVPN domain.

   Note that the GEIs (on PE1 and PE2) of ESI1 don't have to be the
   same.  But if we let them to be the same through configuration, it
   will work well too.

   We use the following encapsulation in this solution:

Wang                     Expires 2 January 2021                [Page 10]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

                +----------------------------------------+
                | IPv4 Header (SIP=VTEP IP)              |
                +----------------------------------------+
                | UDP Header (Source Port ^= ESI label) |
                +----------------------------------------+
                | VXLAN Header (VNI=EGD)                 |
                +----------------------------------------+
                | Ethernet Header                        |
                +----------------------------------------+
                | Ethernet Payload                       |
                +----------------------------------------+
                | Ethernet FCS                           |
                +----------------------------------------+

                Figure 5: VXLAN Encapsulation for EVPN-lite

   Note that only the ingress GEI will be encapsulated in data-plane,
   the VTEP IP of ingress GEI is encapsulated as source IP, the ESI
   label is encrypted into the source port.

   Note that the cryptographic key for ESI label of the inner packet's
   intrinsic entropy.  The intrinsic entropy is a 16 bits unsigned
   integer that is computed from nothing but the inner packet's fields.
   When the ESI label is encrypted into the source port, then the source
   port will contain both the context entropy and the intrinsic entropy
   of the inner packet.  The source port is now called as the
   compositive entropy of the inner packet.

   The GEI/ES route of VXLAN-based EVPN-lite is the RT-1 per ES route.
   The ESI-label attribute is used to carry the GEI1's ESI-label field.
   The OPE TLV or nexthop is used to carry the GEI1's VTEP IP field.

   On receiving the RT-1 per ESI1 route R1 from PE1, PE3 will install a
   GEI mapping entry ME1 into the data-plane.  On receiving a VXLAN
   packet VP1 of Figure 5 format, PE3 recomputes the intrinsic entropy
   of VP1 by the same algorithm, the PE3 decrypts GEI1's ESI label part
   from VP1's UDP source port using the intrinsic entropy.  Then PE3
   learn's VP1's inner S-MAC MAC1 whose destination ESI is ESI1
   according to the GEI mapping entry ME1.

   Note that the simplest encryption algorithm may be the bitwise XOR.
   And it is good enough for our use case.

   On receiving a ethernet packet EP1 whose D-MAC is MAC1, PE3 will
   forwarded EP1 to PE1/PE2 following ESI1's RT-1 per EVI routes for
   EVI1.

Wang                     Expires 2 January 2021                [Page 11]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   On receiving the RT-1 per ESI route R2 from PE1, PE2 will install a
   GEI mapping entry ME2 into the data-plane.  Then, on receiving a
   VXLAN packet VP2 of Figure 5 format, when VP2 is about to be
   forwarded to ESI1, PE2 will drop VP2 because that its ingress GEI is
   ESI1 (according to the GEI mapping entry ME2) too.

3.4.  MPLS Solution Overview

   In MPLS EVPN control plane, we use a 24 bits unsigned number as the
   EGD of EVI1, and it has global uniqueness in EVI1's service domain.
   In data plane, we use QinQ tags to carry the EGD.

   We use a Global Unique Label (GUL) to identify a ESI in EVI1's
   service domain.  So the ESI-GUL is also its Global ESI Indicator.
   The ESI-GULs are avertised through RT-1 per ES routes, and they are
   considered to be an ESI-label by these routes.  The label in RT-3
   route's PMSI-Tunnel Attribute (PTA-Label) whose tunnel type is
   ingress replication is called as Ingress Replication Multicast Label
   (IRML) in this document.

   We use the following encapsulation in MPLS-based EVPN-lite:

            Format #1                    Format #2
       +-----------------------+     +----------------------------+
       | PSN Labels            |     | PSN Labels                 |
       +-----------------------+     +----------------------------+
       | IRML (EVI1)           |     | Destination-ESI GUL (ESI1) |
       +-----------------------+     +----------------------------+
       | Source-ESI GUL (ESI1) |     | Source-ESI GUL (ESI2)      |
       +-----------------------+     +----------------------------+
       | Ethernet Header       |     | Ethernet Header (EVI1)     |
       +-----------------------+     +----------------------------+
       | Ethernet Payload      |     | Ethernet Payload           |
       +-----------------------+     +----------------------------+
       | Ethernet FCS          |     | Ethernet FCS               |
       +-----------------------+     +----------------------------+

                 Figure 6: MPLS Encapsulation for EVPN-lite

   Note that the GUL can be a single Label Stack Entry (LSE), in such
   case, it should be allocated in DCB label space.  Given that the ESIs
   and vESIs may be too many to be allocated in DCB in certain
   scenarios, so the GUL should be allocated in a few context-specific
   label spaces, each identified by a Context Label Space ID (CLS-ID)
   per [I-D.ietf-bess-mvpn-evpn-aggregation-label] in such case.  In
   such case, the ESI-GUL is the entirety of ESI-label and its Context
   Label Space ID (CLS-ID), so it means two LSEs in the Label Stack at
   that time.

Wang                     Expires 2 January 2021                [Page 12]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   Note that the ESI GULs are assigned by a center authority, which may
   be a DC controller or an administrator.

   Note that the ESI-label (ESI-GUL) must be pushed onto the Label Stack
   whether the packet is BUM or not.  The ESI-GUL can't identify the
   EVPN Instance EVI1, so we have to use the EGD in the inner ethernet
   header of "Format #2" to find EVI1 out.

   Note that the GUL concept is very different with the "upstream-
   assigned label (UAL)" concept.  Because that when we receives a GUL
   from a remote PE, the GUL is considered as an outgoing-label to that
   PE, and the GUL is also considered as a incoming-label of the current
   PE, and the label operation for the GUL will be a "swap", to be
   precise, The SPE will swap it to itself and then push the MPLS Label
   Stack to that remote PE.  When the same GUL is received from
   different remote PEs, MPLS ECMP or FRR procedures will be applied.

   So when the GUL is two LSEs in the label stack, we can say that the
   Context-specific Label Space (CLS) of the ESI-label (inside the GUL)
   takes the role of B-VPLS of PBB EVPN, and the CLS-ID label inside the
   GUL takes the role of the B-VPLS label of PBB EVPN.  So no B-VPLS
   instances will be found here.

   Note that the GEI/ES route of MPLS-based EVPN-lite is the RT-1 per ES
   route.

3.5.  SRv6 Solution Overview

   We introduce a SRv6 function named End.ESI to carry the ESI-indicator
   in SRv6 dataplane.  A SID with the End.ESI function is called as an
   "ESI SID" in this document.  The GEI is the locator and fuction part
   of its corresponding ESI SID.  The argument part of the ESI SID is
   the EGD for an EVI.  The EGD works like the function part of an
   End.DT2U/DT2M SID.  But the EGD has a global meaning like a global
   VNI or an PBB ISID but the function part for an End.DT2U/DT2M SID
   typically is only a local discreminator on the egress PE.  The
   argument part of the ESI SID is called as ARG.EGD in this document,
   where the EGD is the abbreviation of EVI-GDV.

   The SRv6 SID in IMET route is an End.DT2M SID with a zero argument
   length.  The GEI1 and GEI2 are SRv6 SID of End.ESI function that is
   defined in the following figure.  We use IGP protocols to advertise
   GEI1 and GEI2 to PE3 respectively in SRv6 underlay.  So we don't use
   EAD/ES route or EAD/EVI route in SRv6 EVPN in this section.  If ESI1
   is single-active mode, GEI1 is different from GEI2, but if ESI1 is
   all-active mode, GEI1 is the same as GEI2.

Wang                     Expires 2 January 2021                [Page 13]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

       |       ESI-Indicator(128-N bits)     |        N bits           |
       +------------+------------+-----------+-------------------------+
       |    Block   |   Node     | ESI.LDV   |        ARG.EGD          |
       +------------+------------+-----------+-------------------------+

                        Figure 7: End.ESI SID Format

   Note that an ESI-indicator is composed of Locator and Function, an
   ESI IP is an 128 bits SID with a zero argument.  The function part is
   a Local Discreminating Value (LDV) on that PE for the ESI.  The
   argument part is a EVI-GDV (EGD) for the EVPN Instance.  The argument
   part is also called ARG.EGD in this document.

3.6.  Comparisons of Relative Concepts

   In this section, we compare the concepts of these four solutions with
   PBB EVPN.  These solutions are:

   *  PBB EVPN: Standard PBB MPLS EVPN per [RFC7623].
   *  Pseudo PBB EVPN: PBB EVPN Control Plane with IP Dataplane, see
      Section 3.2 for details.
   *  NVO3-based EVPN-lite: C-MAC overload Reduction of NVO3 EVPN.
   *  MPLS-based EVPN-lite: C-MAC overload Reduction of MPLS EVPN.
   *  SRv6-based EVPN-lite: C-MAC overload Reduction of SRv6 EVPN.

   We place the detailed comparisons for each solution in separated
   sections, but we place the brief comparisons in the following table:

   +==========+================+=============+============+============+
   | PBB-EVPN |   Pseudo PBB   |  NVO3-based | MPLS-based | SRv6-based |
   +==========+================+=============+============+============+
   | I-VPLS   |     VXLAN      |     EVI     |    EVI     |    EVI     |
   |          |    Instance    |             |            |            |
   +----------+----------------+-------------+------------+------------+
   | I-SID    |      VNI       |     VNI     |    QinQ    |  ARG.EGD   |
   +----------+----------------+-------------+------------+------------+
   | C-MAC    |     C-MAC      |     MAC     |    MAC     |    MAC     |
   +----------+----------------+-------------+------------+------------+
   | B-VPLS   |     IP-VRF     | IP Underlay |   Label    |    SRv6    |
   |          |                |             |   Space    |  Underlay  |
   +----------+----------------+-------------+------------+------------+
   | ESI-BMAC |     ESI IP     | VTEP+S-Port |  ESI-GUL   |  ESI SID   |
   +----------+----------------+-------------+------------+------------+
   | BUM-BMAC |      DIP       |     DIP     |    N/A     |  End.PBB   |
   +----------+----------------+-------------+------------+------------+
   | S-BMAC   |      SIP       |     SIP     | S-ESI GUL  |    SIP     |
   +----------+----------------+-------------+------------+------------+
   | D-BMAC   |      DIP       |     DIP     | D-ESI GUL  |    DIP     |

Wang                     Expires 2 January 2021                [Page 14]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   +----------+----------------+-------------+------------+------------+
   | MPLS     |      MPLS      |    VXLAN    |    MPLS    |    SRv6    |
   | Tunnel   |     Tunnel     |    Tunnel   |   Tunnel   |   Tunnel   |
   +----------+----------------+-------------+------------+------------+

                       Table 1: Concepts Comparisons

   *  I-VPLS - The I-Component of PBB EVPN.
   *  I-SID - The Identifier of I-VPLS, and the EGD of PBB EVPN.
   *  B-VPLS - The B-Component of PBB EVPN
   *  DIP - Destination IP.
   *  SIP - Source IP .
   *  BMAC - Backbone MAC, including Source-BMAC (S-BMAC) and
      Destination-BMAC (D-BMAC).

4.  Dataplane-specific Procedures

4.1.  Packet Walkthrough

   #1 [PE1 forward ARP Request to PE2/PE3]

   *  When CE1 requests CE2's ARP, PE1 will receive the ARP Request BUM1
      from a AC (say AC1) of ESI1.  PE1 will forward the ARP Request
      following the broadcast-list of AC1's EVI instance(say EVI1.  The
      broadcast-list is constructed by IMET routes from PE2/PE3.

      PE1 will forward the ARP Request to PE2/PE3.  The ARP Request is
      encapsulated with GEI1 and EVI1_GDV1.  The inner SMAC of the ARP
      request is M1 which is CE1's MAC address.

   #2  [PE2/PE3's Dataplane MAC Learning]

   *  When PE2/PE3 receives the ARP Request packet BUM1, they do
      dataplane MAC learning independently.  They will learn that M1 is
      behind GEI1.

      Note that when PE2 learns that M1 is behind GEI1, it will assume
      that M1 is behind the local AC whose ESI-indicator is GEI1 too.
      The local AC may have more higher priority than the remote one.

      After the dataplane MAC learning, the ARP request packet BUM1 is
      broadcasted to the local ACs, behind one of which is CE2.

   #3  [PE2 Discard ARP Request to CE1]

Wang                     Expires 2 January 2021                [Page 15]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   *  On receiving BUM1 from PE1, PE2 use the ingress GEI information in
      BUM1 to determine its ingress ESI ESI1, When ESI1 is all-active
      mode and PE2 is about to forward the ARP request to CE1, PE2 will
      find that the ESI for the outgoing AC is also ESI1, so PE2
      discards it for ESI loop-free considerations.

      When ESI1 is single-active mode, the outgoing AC may be in
      blocking state, otherwise its corresponding sub-interface on CE1
      will take charge of packet-drop behavior instead.  So alghough the
      ESI for the outgoing AC is not the same as ESI1, no loop will
      arise in the Ethernet Segment.

   #4  [PE3 Forward ARP Replay to PE1/PE2]

   *  When CE2 replies to CE1 for the ARP request, PE3 will forward the
      ARP reply U1 according to the MAC entry M1 learned previously as
      above.

      PE3 will forward the ARP reply U1 to PE1 or PE2 according to
      ESI1's RT-1 per EVI routes and RT-1 per ES routes:

      When ESI1 is all-active mode, GEI1 may be the same as GEI2, in
      such case, we call both of them GEI21 instead.  The traffics to M1
      will be load-balanced between PE1 and PE2.  Because that GEI21 is
      advertised by both PE1 and PE2l.

   #5  [PE1 Forward ARP Replay to CE1]

   *  Whe PE1 received the ARP reply packet U1 from PE3, PE1 first match
      the packet to the its EVI instance EVI1 by U1's EGD information.
      And PE1 will not discard it because the egress ESI is not the same
      as the ingress ESI which is determined by U1's GEI information.

4.2.  Pseudo PBB-EVPN

   The data plane of Pseudo PBB-EVPN is the same as VXLAN over IP-VRF
   except for a few notable differences.

   Different data packets from PE1 to PE2 will have different SIPs and
   DIPs.  When their egress ESIs are different their DIP will be
   different too.  When their ingress ESIs are different their SIP will
   be different too.  So we can't do <SIP, DIP> (as known as VXLAN
   tunnel) filtering like what have been done in normal VXLAN EVPN
   implements.

Wang                     Expires 2 January 2021                [Page 16]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   Fortunately it is not necessary for us to do such filtering, because
   that the ESI IPs are installed into an IP-VRF not into GRT.  ESI IP
   in the backbone IP-VRF will not have the security issues that normal
   VXLAN EVPNs will have.

   Note that the multicast B-MAC address of BUM packet will be replaced
   with multicast group IP address in the dataplane.  And this requires
   the control-plane to set up the mVPN infrastructures in the backbone
   IP-VRF instance.

   It is a bit weird to use PBB EVPN management plane together with
   VXLAN over IP-VRF data-plane.  So we can use VXLAN over IP-VRF
   management plane instead.

4.3.  VXLAN over IP-VRF

   We configure ESI-IP and VTEP IP instead of B-MAC, VXLAN instance
   instead of I-VPLS, Backbone IP-VRF instead of B-VPLS, VNI instead of
   I-SID.

   The ESI-IP and VTEP IP are advertised by RT-5 routes, not RT-2
   routes.  Although these IPs can be carried in the RT-2 route's IP
   address field too, it may be a bit weird.  So we choose RT-5 route to
   do the advertisement.

   Note that the GEI/ES route in VXLAN over IP-VRF will be RT-5 route,
   And the ESI may be not advertised together with its GEI.

   We don't use multicast IP address as the underlay destination IP
   address of the BUM packets.  We use the VTEP IP address per each
   egress PE instead.  But the IMET route is not advertised for the
   backbone IP-VRF instance, they are advertised for the VXLAN instance.
   We use the Originator Router's IP (ORIP) field of the IMET route as
   the underlay destination IP address instead of the nexthop.  It means
   that ORIP should be advertised for the backbone IP-VRF instance, and
   the ORIP should be the VTEP IP address of corresponding PE.

   There are no other changes from Pseudo PBB-EVPN in data-plane.  So it
   is also a MPLS-based solution.

4.4.  NVO3-specific EVPN-lite Procedures

   In Section 3.3, We use VXLAN encapsulation as an example for NVO3
   EVPN.  But when GENEVE or MPLSoGRE encapsulation is used, the ESI-
   label will have its own space in packet headers, so we don't have to
   encapsulate ESI-label in UDP Source port.

Wang                     Expires 2 January 2021                [Page 17]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   Note that in step #4 the egress GEI is not encapsulated in U1.  U1's
   underlay DIP will be determined by these RT-1 per EVI routes.

4.5.  MPLS-specific EVPN-lite Procedures

   According to [RFC7432], When the IMET route's PTA's tunnel type is
   ingress replication, the ESI-label is considered to be downstream-
   assigned too.  Because that nothing of RT-1 per ES route will
   indicate whether the ESI-label is upstream-assigned or not.

   Alghough ESI-GUL can be a single LSE or two LSEs in the Label Stack,
   we assume that it is a single LSE by default in this section, it is
   for simplification purpose.

   [M1]  In Step #1, "Format #1" of Figure 6 will be used.

         Although the Ingress Replication Multicat Label (IRML) of
         "Format #1" can identify EVI1 by itself, we suppose that the
         ethernet header of it should also carry EGD as what [M4] does.

         Note that there isn't a B-VPLS here, so the IRML identifies the
         EVI1 itself.  The EVI1 here equals C-VPLS of PBB EVPN.

   [M2]  In Step #2, "Format #1" of Figure 6 will be received.  PE3
         knows the packet is for EVI1 with the help of the IRML label.
         Then PE3 can learn the relation between the ingress-GEI
         (ingress-ESI GUL) and S-MAC of BUM1 directly, no GEI to ESI
         lookup needed.

   [M3]  In Step #3, PE2 can compare the ingress-GEI (ingress-ESI GUL)
         of BUM1 and the egress-GEI (ESI-GUL of outgoing AC) directly,
         no GEI to ESI lookup needed.

   [M4]  In Step #4, "Format #2" of Figure 6 will be used.  The source-
         ESI GUL, from which the corresponding MAC entry M1 is
         previously learnt, will be encapsulated as the destination-ESI
         GUL directly.  No GEI to ESI lookup needed only if we don't
         care the requirements of Section 2.9.  Otherwise we should
         refer the corresponding RT-1 per EVI routes of ESI1 to forward
         the packet.  These RT-1 per EVI routes are advertised for EVI1,
         so the Ethernet Tag ID (ETI) of these routes don't have to be
         the EGD.

         Note that when ESI1 is single-active mode, ESI-GUL of ESI1 will
         be different on PE1 and PE2.  But the MAC entry M1 will use the
         newest one only, the swithover between them is called as MAC-
         move.

Wang                     Expires 2 January 2021                [Page 18]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   [M5]  In Step #5, Whe PE1 received the ARP reply packet from PE3, PE1
         first match the packet to ESI1 by Destination-ESI GUL, then
         match the packet to the EVI instance EVI1 by the QinQ tags of
         Ethernet header.

         Note that we suppose that the original tags from ingress AC
         will be processed following the Raw mode per [RFC4448].
         Although the tagged mode can be used technically.  Note that
         the original tags (if they are kept in the packet) will be the
         inner tags of the EGD.

         Note that when RT-1 per EVI route are used, as specified in
         [M4].  There is no need to carry EGD in unicast data-packets
         too.

4.6.  SRv6-specific EVPN-lite Procedures

   [6A]  In Step #1, PE1 will forward the ARP Request to PE2/PE3 with
         the following SRv6 BE encapsulation: It's underlay Source IP is
         the End.ESI SID on PE1 for ESI1; It's underlay Destination IP
         is the End.DT2M SID on PE2/PE3.  The locator and function part
         of the End.ESI SID is GEI1.  The Argument part of the End.ESI
         SID is 0.

         Note that the underlay SIP will be the End.DT2U SID for the
         single-homed ingress ACs.  The multi-homed ingress ACs with
         single-active behavior may not be assigned with an ESI-
         indicator either.  In such situations, the underlay SIP will be
         the End.DT2U SID too.

   [6B]  In Step #3, PE2 can compare the ingress-GEI of BUM1 and the GEI
         of outgoing AC directly, no GEI to ESI lookup needed.

   [6C]  In Step #4, PE3 will forward the ARP reply to PE1 with the
         following SRv6 BE encapsulation: It's underlay Source IP is the
         End.ESI SID on PE3 for ESI2; It's underlay Destination IP is
         the End.ESI SID on PE1 for ESI1 according to the MAC entry M1.
         The ARG.EGD for the End.ESI SID in DIP is the EGD configured on
         PE3.  Note that the EGD for the same EVI is configured with the
         same value on PE1/PE2/PE3.

         When ESI1 is all-active mode, GEI1 will be the same as GEI2, so
         we call both of them GEI21 instead.  The traffics to M1 will be
         load-balanced between PE1 and PE2 by the underlay network on
         PE3.  Because GEI21 is advertised by both PE1 and PE2 in the
         underlay IGP protocol.

Wang                     Expires 2 January 2021                [Page 19]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   [6D]  In Step #5, Whe PE1 received the SRv6 encapsulated ARP reply
         packet from PE3, PE1 first match the packet to the End.ESI SID
         of ESI1 by DIP, then match the packet to the EVI instance EVI1
         by ARG.EGD.

5.  Other Considerations

5.1.  ESI Indicator Advertisement Optimization

   Although we can advertise End.ESI SID in underlay IGP protocols, But
   it is better to use the SRv6 SID Structure Sub-Sub-TLV to indicate
   the length of the ARG.EGD in the End.ESI SID.

   So we can use EAD/EVI route to advertise Global ESI Indicator (GEI),
   these EAD/EVI routes is called as GEI/EVI route in this document.
   When the GEI/EVI route is used to advertise GEI, the End.ESI SID is
   advertised in its SRv6 L2 Service TLV, not in its nexthop.

   Either GEI/EVI routes or GEI/ES routes will be advertised/imported
   for Global Routing Table (GRT), so their Route-Targets (RT) will be
   configured with GRT.  Because there isn't a dedicated B-component
   like PBB VPLS and PBB EVPN.

   Although GEIs is imported to GRT, they are awared only on PE nodes,
   the transit nodes in underlay network won't be aware of GEIs in order
   to reduce the FIB consumption.  We can use the argument length in the
   SRv6 SID Structure Sub-Sub-TLV to check whether the EGD is too big
   for the End.ESI SID, So we can avoid the destruction to the function
   part of the End.ESI and we can use flexible EGD length.

5.2.  C-MAC Flush Notification Procedure

   The withdraw of GEI Advertisement can be used as C-MAC flush
   notification like what have been done by [RFC8317] and
   [I-D.snr-bess-pbb-evpn-isid-cmacflush].

5.3.  E-Tree Support Considerations

   E-tree Supprot extensions is similar to [RFC8317] section 5 except
   for the following notable differences: The leaf B-MACs are replaced
   by leaf GEIs, the root B-MACs are replaced by root GEIs.  the PBB
   encapsulation is replaced by other encapsulations, the B-component is
   replaced by an IP-VRF or the underlay GRT.  The B-MAC Advertisement
   Route is replaced by GEI/EVI route or ESI/IP Route.

Wang                     Expires 2 January 2021                [Page 20]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

5.4.  EVPN IRB Support Considerations

   The dataplane in this draft is no more complex with typical SRv6
   EVPN.  So it will work as efficient as we should expect in SRv6 EVPN
   IRB usecase.

5.5.  Use End.ESI SID in MAC/IP Advertisement Routes

   In [I-D.dawra-bess-srv6-services] the downstream assigned ESI label
   is encapsulated in the Arg.FE2 part of End.DT2M SID, And the ESI
   label present as Arg.FE2 only when the egress PE is adjacent with the
   ingress ESI.  So it is difficult (if not impossible) to do data-plane
   C-MAC learning via End.DT2M SID and its unwarranted Arg.FE2 presence.
   Alghough upstream assigned ESI label may be used to learn ingress
   ESI-indicator on egress PE node, other issues will arise together.

   But the End.ESI SID can be used in MAC/IP advertisement route, only
   if C-MAC overload is not a real threat.  By doing this, the data-
   plane can be unified among these usecases.  The details for using
   End.ESI SID in MAC/IP Advertisement Route will be described in future
   versions.

5.6.  Hierarchical VPLS in EVPN-lite

   In hierachical topology (as illustrated in the following figure), the
   PEs are separated into two groups, the Target PEs (TPEs) and the
   Superstratum PEs (SPEs).

              ___TPE5___        SPE3       ___TPE4_____
             /AC5       \      /   \      /            \AC4
          CE3            \    /     \    /              >=====CE2
             \___         \  /       \  /          ____/AC2
              ___TPE3----SPE1-------SPE2-------TPE2
             /AC3          /                       \
          CE1         ____/                         \
             \____TPE1                               \___CE6
              AC1

                         Figure 8: EVPN-lite H-VPLS

   The TPEs works like the IB-BEB-PE in PBB VPLS, the SPE works like the
   BCB-PE in PBB VPLS.  The BCB-PEs in PBB VPLS do BUM replication based
   on the PBB header.  There are no PBB hearder in EVPN-lite solutions,
   but the SPEs won't learn the C-MACs, which is the same as BCB-PEs in
   PBB VPLS.  The forwarding behaviors of these EVPN-lite solutions are
   very different from each other:

Wang                     Expires 2 January 2021                [Page 21]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   *  The SPE in Pseudo PBB-EVPN do BUM replication based on the
      Multicast Group IP address.

   *  The SPEs in VXLAN over IP-VRF needn't aware of the BUM packets,
      because the destination IP address of the BUM packets will be an
      ingress replication tunnel address to the egress TPE.

   *  The SPEs in MPLS-based EVPN-lite don't have to aware of the BUM
      packets, because that, for IMET routes, they work like the ASBRs
      in inter-AS option B.  In such case, the TPEs do ingress-
      replication for all other TPEs by themselves.

      The SPEs in MPLS-based EVPN-lite may terminate the IMET routes
      that were received from their TPEs.  These IMET routes are
      imported into an corresponding BD, but may not be passed through
      other SPEs, so as not to cause duplicated BUM packets.  In such
      case, take SPE1 for example, there are two split-horizon-groups,
      one group is TPE1/TPE3/TPE5, another split-horizon-group is SPE1/
      SPE2.  The BUM packets are replicated between different split-
      horizon-groups.  In such case, the TPEs do ingress-replication for
      its directly connected TPEs and SPEs, not for the indirectly
      connected TPEs and SPEs.  But the unicast packet will not be
      forwarded by that BD on the SPEs.  The unicast packets will be
      label-swapped in the context-specific label-space for the
      corresponding GULs.

      Note that the BCB-PE in PBB VPLS is typically supported in the
      industry, But it seems that the BCB-PE in PBB EVPN is typically
      not supported in the industry up to now.  Because the BCB-PE
      function can be replaced in MPLS EVPN by a label-swapping
      operation which is like the inter-AS option B scenarios.

   Note that the BUM packets here are defined based on the destination
   C-MAC addresses.

6.  Comparison with Other Solutions

6.1.  Questions

   We compare EVPN-lite with three other solutions in this section.
   These solutions are:

   *  PBB EVPN - PBB MPLS EVPN per [RFC7623].
   *  PBB VPLS - PBB VPLS (PW-based) per [RFC7041].
   *  EVPN VXLAN - [RFC7348] plus IMET routes of [RFC8365], the IMET
      routes are used to discover EVIs and establish VXLAN tunnels.

Wang                     Expires 2 January 2021                [Page 22]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   We use the following questions for these solutions to do the
   comparison:

   [CMAC-Reduction]
   #1  No C-MAC Awareness in the Backbone ?

   [EVPN-IRB]
   #2  EVPN IRB Support ?

   [Unified-Encap]
   #3  Unified Encapsulation per Scenario ?

   [ESI-Retained]
   #4  ESI Features Remain Supported ?

   [Flexible-MH]
   #5  Flexible Multi-homing Remains Supported ?

   [Learn-Confined]
   #6  C-MAC Address Learning and Confinement ?

   [AA-no-flush]
   #7  No C-MAC Flushing for All-Active ESes ?

   [SA-independent]
   #8  Independent C-MAC Flushing for Single-Active ESes ?

   [<ESI,EVI> Converge]
   #9  Independent Convergency per <ESI, EVI> ?

   [Route Aggregation]
   #10 Route Aggregation and Default Route in Backbone ?

6.2.  Summary Comparisons

   We place the detailed comparisons about the answers of these
   questions for each solution in separated sections, but we place the
   brief comparisons in the following table:

   +===================+===========+==========+==========+============+
   | Questions         | EVPN-lite | PBB-EVPN | PBB-VPLS | EVPN-VXLAN |
   +===================+===========+==========+==========+============+
   | CMAC-Reduction    |    Yes    |   Yes    |   Yes    |    Yes     |
   +-------------------+-----------+----------+----------+------------+
   | EVPN-IRB          |    Yes    |    No    |    No    |    Yes     |
   +-------------------+-----------+----------+----------+------------+
   | Unified-Encap     |    Yes    |    No    |    No    |    Yes     |
   +-------------------+-----------+----------+----------+------------+

Wang                     Expires 2 January 2021                [Page 23]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   | ESI-Retained      |    Yes    |   Yes    |    No    |     No     |
   +-------------------+-----------+----------+----------+------------+
   | Flexible-MH       |    Yes    |   Yes    |   N/A    |     No     |
   +-------------------+-----------+----------+----------+------------+
   | Learn-Confined    |    Yes    |   Yes    |   Yes    |    Yes     |
   +-------------------+-----------+----------+----------+------------+
   | AA-no-flush       |    Yes    |   Yes    |    No    |    Yes     |
   +-------------------+-----------+----------+----------+------------+
   | SA-independent    |    Yes    |   Yes    |    No    |     No     |
   +-------------------+-----------+----------+----------+------------+
   | ESI-EVI-Converge  |    Yes    |    No    |    No    |     No     |
   +-------------------+-----------+----------+----------+------------+
   | Route-Aggregation |    Yes    |    No    |    NO    |    N/A     |
   +-------------------+-----------+----------+----------+------------+

                      Table 2: Solution Comparisons

6.3.  Detailed Comparisons with PBB EVPN

   TBD.

6.4.  Detailed Comparisons with PBB VPLS

   TBD.

6.5.  Detailed Comparisons with EVPN VXLAN

   TBD.

7.  Security Considerations

   Detailed in Section 4.2.  Other considerations will be added in
   future versions.

8.  IANA Considerations

8.1.  END.ESI SID

   IANA is requested to allocate a new code points for the new SRv6
   Endpoint Behaviors defined in this document.

                  +------+-------------+---------------+
                  | Type | Description | Reference     |
                  +------+-------------+---------------+
                  | TBD1 | END.ESI     | This Document |
                  +------+-------------+---------------+

Wang                     Expires 2 January 2021                [Page 24]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

                             Figure 9: END.ESI

8.2.  Global Unique ESI-label in EAD per ES Route

   When we use Global Unique ESI-label in EAD per ES route, especially
   in ingress-replication use case, It should be explicitly indicated in
   the EAD per ES route.  The details will be added in future versions.

9.  Acknowledgements

   The authors would like to thank the following for their comments and
   review of this document:

   Ye Shu.

10.  References

10.1.  Normative References

   [I-D.dawra-bess-srv6-services]
              Dawra, G., Filsfils, C., Brissette, P., Agrawal, S.,
              Leddy, J., daniel.voyer@bell.ca, d.,
              daniel.bernier@bell.ca, d., Steinberg, D., Raszuk, R.,
              Decraene, B., Matsushima, S., Zhuang, S., and J. Rabadan,
              "SRv6 BGP based Overlay services", Work in Progress,
              Internet-Draft, draft-dawra-bess-srv6-services-02, 8 July
              2019, <https://tools.ietf.org/html/draft-dawra-bess-srv6-
              services-02>.

   [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", Work in
              Progress, Internet-Draft, draft-ietf-bess-mvpn-evpn-
              aggregation-label-03, 24 October 2019,
              <https://tools.ietf.org/html/draft-ietf-bess-mvpn-evpn-
              aggregation-label-03>.

   [RFC4448]  Martini, L., Ed., Rosen, E., El-Aawar, N., and G. Heron,
              "Encapsulation Methods for Transport of Ethernet over MPLS
              Networks", RFC 4448, DOI 10.17487/RFC4448, April 2006,
              <https://www.rfc-editor.org/info/rfc4448>.

   [RFC7348]  Mahalingam, M., Dutt, D., Duda, K., Agarwal, P., Kreeger,
              L., Sridhar, T., Bursell, M., and C. Wright, "Virtual
              eXtensible Local Area Network (VXLAN): A Framework for
              Overlaying Virtualized Layer 2 Networks over Layer 3
              Networks", RFC 7348, DOI 10.17487/RFC7348, August 2014,
              <https://www.rfc-editor.org/info/rfc7348>.

Wang                     Expires 2 January 2021                [Page 25]
Internet-Draft            EVPN C-MAC Reduction                 July 2020

   [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, <https://www.rfc-editor.org/info/rfc7432>.

   [RFC7623]  Sajassi, A., Ed., Salam, S., Bitar, N., Isaac, A., and W.
              Henderickx, "Provider Backbone Bridging Combined with
              Ethernet VPN (PBB-EVPN)", RFC 7623, DOI 10.17487/RFC7623,
              September 2015, <https://www.rfc-editor.org/info/rfc7623>.

   [RFC8317]  Sajassi, A., Ed., Salam, S., Drake, J., Uttaro, J.,
              Boutros, S., and J. Rabadan, "Ethernet-Tree (E-Tree)
              Support in Ethernet VPN (EVPN) and Provider Backbone
              Bridging EVPN (PBB-EVPN)", RFC 8317, DOI 10.17487/RFC8317,
              January 2018, <https://www.rfc-editor.org/info/rfc8317>.

   [RFC8365]  Sajassi, A., Ed., Drake, J., Ed., Bitar, N., Shekhar, R.,
              Uttaro, J., and W. Henderickx, "A Network Virtualization
              Overlay Solution Using Ethernet VPN (EVPN)", RFC 8365,
              DOI 10.17487/RFC8365, March 2018,
              <https://www.rfc-editor.org/info/rfc8365>.

10.2.  Informative References

   [I-D.snr-bess-pbb-evpn-isid-cmacflush]
              Rabadan, J., Sathappan, S., Nagaraj, K., Miyake, M., and
              T. Matsuda, "PBB-EVPN ISID-based CMAC-Flush", Work in
              Progress, Internet-Draft, draft-snr-bess-pbb-evpn-isid-
              cmacflush-06, 26 July 2019, <https://tools.ietf.org/html/
              draft-snr-bess-pbb-evpn-isid-cmacflush-06>.

   [RFC7041]  Balus, F., Ed., Sajassi, A., Ed., and N. Bitar, Ed.,
              "Extensions to the Virtual Private LAN Service (VPLS)
              Provider Edge (PE) Model for Provider Backbone Bridging",
              RFC 7041, DOI 10.17487/RFC7041, November 2013,
              <https://www.rfc-editor.org/info/rfc7041>.

Author's Address

   Yubao Wang
   ZTE Corporation
   No. 50 Software Ave, Yuhuatai Distinct
   Nanjing
   China

   Email: yubao.wang2008@hotmail.com

Wang                     Expires 2 January 2021                [Page 26]