Diff: rfc9251xml2.original.xml

	rfc9251xml2.original.xml	rfc9251.xml

	<?xml version='1.0'?>	<?xml version='1.0' encoding='utf-8'?>
	<!DOCTYPE rfc SYSTEM 'rfc2629.dtd' [	<!DOCTYPE rfc [
	]>	<!ENTITY nbsp " ">
	<?rfc toc="yes"?>	<!ENTITY zwsp "">
	<?rfc tocompact="no"?>	<!ENTITY nbhy "‑">
	<?rfc tocdepth="6"?>	<!ENTITY wj "⁠">
	<?rfc symrefs="yes"?>	]>
	<?rfc sortrefs="yes"?>
	<?rfc compact="yes"?>
	<?rfc subcompact="no"?>
	<?rfc strict="yes" ?>
	<rfc category="std" docName="draft-ietf-bess-evpn-igmp-mld-proxy-21">
	<!-- *** FRONT MATTER *** -->
	<front>
	<title abbrev="IGMP and MLD Proxy for EVPN">IGMP and MLD Proxy for EVPN</tit
	le>


	<author initials="A" surname="Sajassi"	<rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="std" docName="draft-i
	fullname="Ali Sajassi">	etf-bess-evpn-igmp-mld-proxy-21" obsoletes="" updates="" submissionType="IETF" c
	<organization>Cisco Systems</organization>	onsensus="true" ipr="trust200902" xml:lang="en" tocInclude="true" tocDepth="6" s
		ymRefs="true" sortRefs="true" version="3" number="9251">
		<!-- xml2rfc v2v3 conversion 3.12.2 -->


	<address>	<front>
	<postal>
	<street>821 Alder Drive,</street>


	<region>MILPITAS, CALIFORNIA 95035</region>	<!--[rfced] Please note that the title of the document has been
		updated as follows. Abbreviations have been expanded per Section
		3.6 of RFC 7322 ("RFC Style Guide") and "Proxy" has been updated
		to "Proxies". Please review.


	<country>UNITED STATES</country>	Original:
	</postal>	IGMP and MLD Proxy for EVPN


	<phone></phone>	Current:
	<email>sajassi@cisco.com</email>	Internet Group Management Protocol (IGMP) and
	</address>	Multicast Listener Discovery (MLD) Proxies for Ethernet VPN (EVPN)
		-->

		<title abbrev="IGMP and MLD Proxies for EVPN"> Internet Group Management Pro
		tocol (IGMP) and Multicast Listener Discovery (MLD) Proxies for Ethernet VPN (EV
		PN)</title>
		<seriesInfo name="RFC" value="9251"/>
		<author initials="A" surname="Sajassi" fullname="Ali Sajassi">
		<organization>Cisco Systems</organization>
		<address>
		<postal>
		<street>821 Alder Drive</street>
		<code>95035</code>
		<city>Milpitas</city>
		<region>CA</region>
		<country>United States of America</country>
		</postal>
		<phone/>
		<email>sajassi@cisco.com</email>
		</address>
	</author>	</author>

	<author initials="S" surname="Thoria"	<author initials="S" surname="Thoria" fullname="Samir Thoria">
	fullname="Samir Thoria">
	<organization>Cisco Systems</organization>	<organization>Cisco Systems</organization>

	<address>	<address>
	<postal>	<postal>
	<street>821 Alder Drive,</street>	<street>821 Alder Drive</street>
		<code>95035</code>
	<region>MILPITAS, CALIFORNIA 95035</region>	<city>Milpitas</city>
		<region>CA</region>
	<country>UNITED STATES</country>	<country>United States of America</country>
	</postal>	</postal>
		<phone/>
	<phone></phone>	<email>sthoria@cisco.com</email>
	<email>sthoria@cisco.com</email>	</address>
	</address>
	</author>	</author>

	<author initials="M" surname="Mishra"	<author initials="M" surname="Mishra" fullname="Mankamana Mishra">
	fullname="Mankamana Mishra">
	<organization>Cisco Systems</organization>	<organization>Cisco Systems</organization>

	<address>	<address>
	<postal>	<postal>
	<street>821 Alder Drive,</street>	<street>821 Alder Drive</street>
		<code>95035</code>
	<region>MILPITAS, CALIFORNIA 95035</region>	<city>Milpitas</city>
		<region>CA</region>
	<country>UNITED STATES</country>	<country>United States of America</country>
	</postal>	</postal>
		<phone/>
	<phone></phone>	<email>mankamis@cisco.com</email>
	<email>mankamis@cisco.com</email>	</address>
	</address>
	</author>	</author>

		<author initials="K" surname="Patel" fullname="Keyur Patel">
	<author initials="K" surname="Patel"
	fullname="Keyur PAtel">
	<organization>Arrcus</organization>	<organization>Arrcus</organization>

	<address>	<address>
	<postal>	<postal>
	<street></street>	<country>United States of America</country>
		</postal>
	<region></region>	<phone/>
		<email>keyur@arrcus.com</email>
	<country>UNITED STATES</country>	</address>
	</postal>

	<phone></phone>
	<email>keyur@arrcus.com</email>
	</address>
	</author>	</author>

		<author initials="J" surname="Drake" fullname="John Drake">
	<author initials="J" surname="Drake"
	fullname="John Drake">
	<organization>Juniper Networks</organization>	<organization>Juniper Networks</organization>

	<address>	<address>
	<postal>	<email>jdrake@juniper.net</email>
	<street></street>	</address>

	<region></region>

	<country></country>
	</postal>

	<phone></phone>
	<email>jdrake@juniper.net</email>
	</address>
	</author>	</author>

		<author initials="W" surname="Lin" fullname="Wen Lin">
	<author initials="W" surname="Lin"
	fullname="Wen Lin">
	<organization>Juniper Networks</organization>	<organization>Juniper Networks</organization>

	<address>	<address>
	<postal>	<email>wlin@juniper.net</email>
	<street></street>	</address>
		</author>
		<date year="2022" month="May"/>
		<area>RTG</area>
		<workgroup>BESS</workgroup>


	<region></region>	<!-- [rfced] Please insert any keywords (beyond those that appear in
		the title) for use on https://www.rfc-editor.org/search. -->


	<country></country>	<abstract>
	</postal>	<!--[rfced] In this sentence, is "efficiently" meant to describe "support"
		or "running"?


	<phone></phone>	Original:
	<email>wlin@juniper.net</email>	This document describes how to support efficiently endpoints running
	</address>	IGMP(Internet Group Management Protocol) or MLD (Multicast Listener
	</author>	Discovery)...


	<date year="2022"/>	Perhaps (describing "support"):
	<area>Routing</area>	This document describes how to efficiently support endpoints running
	<workgroup>BESS WorkGroup</workgroup>	Internet Group Management Protocol (IGMP) or Multicast Listener
	<abstract>	Discovery (MLD)...


	<t>	Or (describing "running"):
	This document describes how to support efficiently endpoints runn	This document describes how to support endpoints efficiently running
	ing	Internet Group Management Protocol (IGMP) or Multicast Listener
	IGMP(Internet Group Management Protocol) or MLD (Multicast Listen	Discovery (MLD)...
	er Discovery) for the multicast services	-->
	over an EVPN network by incorporating
	IGMP/MLD proxy procedures on EVPN (Ethernet VPN) PEs.	<t>This document describes how to support efficiently endpoints running
	</t>	the Internet Group Management Protocol (IGMP) or Multicast Listener Discov
		ery (MLD)
		for the multicast services over an Ethernet VPN (EVPN) network by incorpor
		ating
		IGMP/MLD proxy procedures on EVPN Provider Edges (PEs).
		</t>
	</abstract>	</abstract>
	</front>	</front>


	<!-- *** MIDDLE MATTER *** -->

	<middle>	<middle>

	<section title="Introduction">	<section numbered="true" toc="default">
		<name>Introduction</name>
	<t>	<t>In data center (DC) applications, a point of delivery (POD) can consist
	In DC applications, a point of delivery (POD) can consist of	of a
	a	collection of servers supported by several top-of-rack (ToR) and
	collection of servers supported by several top of rack (ToR)	spine switches. This collection of servers and switches are self-contained
	and	and may have their own control protocol for intra-POD
	spine switches. This collection of servers and switches are	communication and orchestration. However, EVPN is used as a standard
	self	way of inter-POD communication for both intra-DC and inter-DC. A
	contained and may have their own control protocol for intra-	subnet can span across multiple PODs and DCs. EVPN provides a robust
	POD	multi-tenant solution with extensive multihoming capabilities to
	communication and orchestration. However, EVPN is used as st	stretch a subnet (VLAN) across multiple PODs and DCs. There can be
	andard	many hosts (several hundreds) attached to a subnet that is
	way of inter-POD communication for both intra-DC and inter-D	stretched across several PODs and DCs.
	C. A	</t>
	subnet can span across multiple PODs and DCs. EVPN provides	<t>These hosts express their interests in multicast groups on a
	a robust	given subnet/VLAN by sending IGMP/MLD Membership Reports for
	multi-tenant solution with extensive multi-homing capabiliti	their interested multicast group(s). Furthermore, an IGMP/MLD router
	es to	periodically sends membership queries to find out if there are hosts
	stretch a subnet (VLAN) across multiple PODs and DCs. There	on that subnet that are still interested in receiving multicast
	can be	traffic for that group. The IGMP/MLD Proxy solution described in this
	many hosts (several hundreds) attached to a subnet that is	document accomplishes three objectives:
	stretched across several PODs and DCs.	</t>
		<ol spacing="normal" type="1">
	</t>	<li>Reduce flooding of IGMP/MLD messages: Just like the ARP / Neighbor Di
		scovery (ND)
	<t>	suppression
	These hosts express their interests in multicast groups	mechanism in EVPN to reduce the flooding of ARP messages over EVPN,
	on a	it is also desired to have a mechanism to reduce the flooding of IGMP/MLD
	given subnet/VLAN by sending IGMP/MLD Membership Reports	messages (both Queries and Membership Reports) in EVPN.</li>
	for	<li>Distributed anycast multicast proxy: It is desirable for the EVPN
	their interested multicast group(s). Furthermore, an IGM	network to act as a distributed anycast multicast router with respect
	P/MLD router	to IGMP/MLD proxy function for all the hosts attached to that
	periodically sends membership queries to find out if the	subnet.</li>
	re are hosts	<li>Selective multicast: This describes forwarding multicast traffic ove
	on that subnet that are still interested in receiving mu	r the EVPN
	lticast	network such that it only gets forwarded to the PEs that have
	traffic for that group. The IGMP/MLD Proxy solution desc	interests in the multicast group(s). This document shows how this objecti
	ribed in this	ve may be achieved
	document accomplishes three objectives:	when ingress replication is used to distribute the multicast traffic
		among the PEs. Procedures for supporting selective multicast using
	<list style="numbers">	Point-to-Multipoint (P2MP) tunnels can be found in <xref target="I-D.ietf
	<t>	-bess-evpn-bum-procedure-updates"
	Reduce flooding of IGMP/MLD messages: ju	format="default"/>.</li>
	st like the ARP/ND suppression	</ol>
	mechanism in EVPN to reduce the flooding	<t>The first two objectives are achieved by using the IGMP/MLD proxy on th
	of ARP messages over EVPN,	e
	it is also desired to have a mechanism t	PE. The third objective is achieved by setting up a multicast
	o reduce the flooding of IGMP/MLD	tunnel among only the PEs that have
	messages (both Queries and Membership Re	interest in the multicast group(s) based on the trigger from
	ports) in EVPN.	IGMP/MLD proxy processes. The proposed solutions for each of these
	</t>	objectives are discussed in the following sections.
		</t>
	<t>	</section>
	Distributed anycast multicast proxy	<section numbered="true" toc="default">
	: it is desirable for the EVPN	<name>Specification of Requirements</name>
	network to act as a distributed any	<t>The key words "<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>",
	cast multicast router with respect	"<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL NOT</bcp1
	to IGMP/MLD proxy function for all	4>",
	the hosts attached to that	"<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED<
	subnet.	/bcp14>",
	</t>	"<bcp14>NOT RECOMMENDED</bcp14>", "<bcp14>MAY</bcp14>", and
		"<bcp14>OPTIONAL</bcp14>" in this document are to be interpreted as descri
	<t>	bed in
	Selective Multicast: to forward	BCP 14 <xref target="RFC2119" format="default"/> <xref target="RFC8174"
	multicast traffic over EVPN	format="default"/> when, and only when, they appear in all capitals, as s
	network such that it only gets f	hown here.
	orwarded to the PEs that have	</t>
	interest in the multicast group(	</section>
	s). This document shows how this objective may be achieved	<section numbered="true" toc="default">
	when Ingress Replication is used	<name>Terminology</name>
	to distribute the multicast traffic	<dl newline="false" spacing="normal">
	among the PEs. Procedures for s	<dt> AC:</dt>
	upporting selective multicast using	<dd> Attachment Circuit</dd>
	P2MP tunnels can be found in <xr	<dt>All-Active Redundancy Mode:</dt>
	ef target="I-D.ietf-bess-evpn-bum-procedure-updates"/>	<dd> When all PEs attached to an Ethernet
	</t>	segment are allowed to forward known unicast traffic to/from that
	</list>	Ethernet segment for a given VLAN, then the Ethernet segment is
	</t>	defined to be operating in All-Active redundancy mode.</dd>
		<dt>BD:</dt>
		<dd> Broadcast Domain. As per <xref target="RFC7432" format="default"/>,
		an EVPN instance
		(EVI) consists of a single BD
		or multiple BDs. In case of a VLAN bundle and a VLAN-aware bundle service
		model, an EVI contains multiple BDs. Also, in this document, BD and
		subnet are equivalent terms.</dd>
		<dt>DC:</dt>
		<dd> Data Center</dd>
		<dt>ES:</dt>
		<dd> Ethernet Segment. This is when a customer site (device or network) i
		s
		connected to one or more PEs via a set of Ethernet links.</dd>
		<dt>ESI:</dt>
		<dd> Ethernet Segment Identifier. This is a unique non-zero identifier th
		at
		identifies an Ethernet Segment.</dd>
		<dt>Ethernet Tag:</dt>
		<dd> It identifies a particular broadcast
		domain, e.g., a VLAN. An EVPN instance consists of one or more
		broadcast domains.</dd>
		<dt>EVI:</dt>
		<dd> EVPN Instance. This spans the Provider Edge (PE) devices
		participating in that EVPN.</dd>
		<dt>EVPN:</dt>
		<dd> Ethernet Virtual Private Network</dd>
		<dt>IGMP:</dt>
		<dd> Internet Group Management Protocol</dd>
		<dt>IR:</dt>
		<dd> Ingress Replication</dd>
		<dt>MLD:</dt>
		<dd> Multicast Listener Discovery</dd>
		<dt> OIF:</dt>
		<dd> Outgoing Interface for multicast. It can be a physical interface,
		virtual interface, or tunnel.</dd>
		<dt>PE:</dt>
		<dd> Provider Edge</dd>
		<dt>POD:</dt><dd> Point of Delivery</dd>
		<dt> S-PMSI:</dt>
		<dd> Selective P-Multicast Service Interface. This is a conceptual interf
		ace for a
		PE to send customer multicast traffic to some of the PEs in the same VPN.
		</dd>
		<dt>Single-Active Redundancy Mode:</dt>
		<dd> When only a single PE, among all the
		PEs attached to an Ethernet segment, is allowed to forward traffic
		to/from that Ethernet segment for a given VLAN, then the Ethernet
		segment is defined to be operating in Single-Active redundancy mode.</dd>
		<dt> SMET:</dt>
		<dd> Selective Multicast Ethernet Tag</dd>
		<dt>ToR:</dt>
		<dd> Top of Rack</dd>
		</dl>
		<t>This document also assumes familiarity with the terminology of
		<xref target="RFC7432" format="default"/>, <xref target="RFC3376"
		format="default"/>, and <xref target="RFC2236" format="default"/>.


	<t>	<!--[rfced] May we rephrase this sentence for clarity? Please let
	The first two objectives are achieved by using IGMP/MLD	us know if the suggested text is agreeable or if you prefer
	proxy on the	otherwise.
	PE. The third objective is achieved by setting up a mul
	ticast
	tunnel only among the PEs that have
	interest in that multicast group(s) based on the trigge
	r from
	IGMP/MLD proxy processes. The proposed solutions for ea
	ch of these
	objectives are discussed in the following sections.
	</t>
	</section>


	<section title="Specification of Requirements">	Original:
	<t>	Though most of the place this
	The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "S	document uses term IGMP Membership Report, the text applies equally
	HALL NOT",	for MLD Membership Report too.
	"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED",
	"MAY", and
	"OPTIONAL" in this document are to be interpreted as desc
	ribed in BCP
	14 <xref target="RFC2119"/> <xref target="RFC8174"/> when
	, and only when,
	they appear in all
	capitals, as shown here.
	</t>


	</section>	Perhaps:
		When this document uses the term "IGMP Membership Report", the text
		equally applies to the MLD Membership Report.
		-->


	<section title="Terminology">	Though most
	<t>	of the place this document uses the term "IGMP
	<list style="symbols">	Membership Report", the text applies equally for MLD
	<t> AC: Attachment Circuit.	Membership Report too. Similarly, text for IGMPv2 applies to MLDv1,
	</t>	and text for IGMPv3 applies to MLDv2. IGMP/MLD version encoding in the
	<t>	BGP update is stated in <xref target="bgp-encoding" format="default"/>.</t
	All-Active Redundancy Mode: When all PEs attached to a	>
	n Ethernet	<t> It is important to note that when there is text considering whether a
	segment are allowed to forward known unicast traffic t	PE
	o/from that	indicates support for IGMP proxying, the corresponding behavior has a
	Ethernet segment for a given VLAN, then the Ethernet s	natural analog for indicating support for MLD proxying, and the analogous
	egment is	requirements apply as well.
	defined to be operating in All-Active redundancy mode.	</t>
	</t>	</section>
	<t>	<section numbered="true" toc="default">
	BD: Broadcast Domain. As per <xref target="RFC7432"/>,	<name>IGMP/MLD Proxy</name>
	an EVI consists of a single	<t>The IGMP Proxy mechanism is used to reduce the flooding of IGMP
	or multiple BDs. In case of VLAN-bundle and VLAN-aware	messages over an EVPN network, similar to the ARP proxy used in reducing
	bundle service	the flooding of ARP messages over EVPN. It also provides a triggering
	model, an EVI contains multiple BDs. Also, in this doc	mechanism for the PEs to set up their underlay multicast tunnels. The
	ument, BD and	IGMP Proxy mechanism consists of two components:
	subnet are equivalent terms.	</t>
	</t>	<ol spacing="normal" type="1">
	<t>	<li> Proxy for IGMP Membership Reports </li>
	DC: Data Center	<li> Proxy for IGMP Membership Queries </li>
	</t>	</ol>
		<t>The goal of IGMP and MLD proxying is to make the EVPN behave seamlessly
		for
		the tenant systems with respect to multicast operations while using a more
		efficient delivery system for signaling and delivery across the VPN.
		Accordingly, group state must be tracked synchronously among the PEs
		serving the VPN, with join and leave events propagated to the peer PEs and
		each PE tracking the state of each of its peer PEs with respect to whether
		there are locally attached group members (and in some cases, senders), wha
		t
		version(s) of IGMP/MLD are in use for those locally attached group members
		,
		etc. In order to perform this translation, each PE acts as an IGMP router
		for the locally attached domain, maintains the requisite state on
		locally attached nodes, sends periodic membership queries, etc. The role
		of EVPN Selective Multicast Ethernet Tag (SMET) route propagation is to
		ensure that each PE's local state is
		propagated to the other PEs so that they share a consistent view of the
		overall IGMP Membership Request and Leave Group state. It is important to
		note that the need to keep such local state can be triggered by either
		local IGMP traffic or BGP EVPN signaling. In most cases, a local IGMP eve
		nt
		will need to be signaled over EVPN, though state initiated by received EVP
		N
		traffic will not always need to be relayed to the locally attached domain.
		</t>
		<section numbered="true" toc="default">
		<name>Proxy Reporting</name>


	<t>	<!--[rfced] Should instances of "IGMP protocol" be updated to simply read "IGMP"
	Ethernet Segment (ES): When a customer site (device or	to avoid redundancy (if expanded "IGMP protocol" would read as "Internet Group
	network) is	Management Protocol protocol"). Please review and let us know if any updates
	connected to one or more PEs via a set of Ethernet lin	are needed.
	ks.	-->
	</t>


	<t>	<t>When IGMP protocol is used between hosts and their first hop EVPN
	Ethernet Segment Identifier (ESI): A unique non-zero i	router (EVPN PE), Proxy-reporting is used by the EVPN PE to summarize
	dentifier that	(when possible) reports received from downstream hosts and propagate
	identifies an Ethernet Segment.	them in BGP to other PEs that are interested in the information.
	</t>


	<t>	<!--[rfced] May we update "IGMP Reports" to "IGMP Membership Reports"?
	Ethernet Tag: It identifies a particular broadcast
	domain, e.g., a VLAN. An EVPN instance consists of on
	e or more
	broadcast domains.
	</t>


	<t>	Original:
	EVI: An EVPN instance spanning the Provider Edge (PE)	This is done by terminating the IGMP Reports in the first hop PE, and
	devices	translating and exchanging the relevant information among EVPN BGP
	participating in that EVPN	speakers.
	</t>


	<t>	Perhaps:
	EVPN: Ethernet Virtual Private Network	This is done by terminating the IGMP Membership Reports in the first hop PE,
	</t>	and
		translating and exchanging the relevant information among EVPN BGP
		speakers.
		-->
		This
		is done by terminating the IGMP Reports in the first hop PE and
		translating and exchanging the relevant information among EVPN BGP
		speakers. The information is again translated back to an IGMP message at
		the recipient EVPN speaker. Thus, it helps create an IGMP overlay
		subnet using BGP. In order to facilitate such an overlay, this
		document also defines a new EVPN route type Network Layer Reachability In
		formation
		(NLRI) and the EVPN SMET route, along with its procedures to help
		exchange and register IGMP multicast groups; see <xref target="bgp-encodi
		ng"
		format="default"/>.
		</t>
		<section numbered="true" toc="default">
		<name>IGMP/MLD Membership Report Advertisement in BGP</name>
		<t>When a PE wants to advertise an IGMP Membership Report using
		the BGP EVPN route, it follows the proceeding rules (BGP encoding
		is stated in <xref target="bgp-encoding" format="default"/>). The first
		four
		rules are applicable to the originator PE, and the last three rules are
		applicable
		to remote PE processing SMET routes:
		</t>
		<t>Processing at the BGP route originator:
		</t>
		<ol spacing="normal" type="1">
		<li>When the first hop PE receives IGMP Membership Reports
		belonging to the same IGMP version from different attached
		hosts for the same (*,G) or (S,G), it <bcp14>SHOULD</bcp14> send a si
		ngle
		BGP message corresponding to the very first IGMP Membership Request (
		BGP update as
		soon as possible) for that (*,G) or (S,G). This is because BGP is a
		stateful protocol, and no further transmission of the same report is
		needed. If the IGMP Membership Request is for (*,G), then the Multica
		st Group Address
		<bcp14>MUST</bcp14> be sent along with the corresponding version flag
		(v2 or v3)
		set. In case of IGMPv3, the exclude flag <bcp14>MUST</bcp14> also be
		set to
		indicate that no source IP address must be excluded (include all
		sources "*").
		If the IGMP Membership Report is for (S,G), then besides setting the
		Multicast Group
		Address along with the v3 flag, the source IP address and the
		include/exclude (IE) flag <bcp14>MUST</bcp14> be set. It should be no
		ted that, when
		advertising the EVPN route for (S,G), the only valid version flag is
		v3 (v2 flags <bcp14>MUST</bcp14> be set to 0).
		</li>
		<li>When the first hop PE receives an IGMPv3 Membership Report for (
		S,G) on a given
		BD, it <bcp14>MUST</bcp14> advertise the corresponding EVPN SMET rout
		e, regardless
		of whether the source (S) is
		attached to itself or not, in order to facilitate the source move in
		the future. </li>
		<li>When the first hop PE receives an IGMP version-X Membership Repo
		rt first for
		(*,G) and then later receives an IGMP version-Y Membership Report for
		the same
		(*,G), then it <bcp14>MUST</bcp14> re-advertise the same EVPN SMET ro
		ute with the flag
		for version-Y set in addition to any previously set version flag(s).
		In other words, the first hop PE <bcp14>MUST NOT</bcp14> withdraw the
		EVPN route
		before sending the new route because the Flags field is not part of
		BGP route key processing.
		</li>
		<li>When the first hop PE receives an IGMP version-X Membership Repo
		rt first for
		(*,G) and then later receives an IGMPv3 Membership Report for the sam
		e
		Multicast Group Address but for a specific source address S, then the
		PE <bcp14>MUST</bcp14> advertise a new EVPN SMET route with the v3 fl
		ag set (and v2 reset).
		The IE flag also needs to be set accordingly.
		Since the source IP address is used as part of BGP route key processi
		ng,
		it is considered to be a new BGP route advertisement. When different
		versions
		of IGMP Membership Report are received, the final state <bcp14>MUST</
		bcp14> be as per
		<xref target="RFC3376" sectionFormat="of" section="5.1"/>.
		At the end of the route processing, local and remote group record sta
		te
		<bcp14>MUST</bcp14>
		be as per <xref target="RFC3376" sectionFormat="of" section="5.1"/>.
		</li>
		</ol>
		<t>Processing at the BGP route receiver:
		</t>
		<ol spacing="normal" type="1">
		<li>When a PE receives an EVPN SMET route with more than one version
		flag set, it will generate the corresponding IGMP report for (*,G)
		for each version specified in the Flags field. With multiple version
		flags set, there must not be a source IP address in the received EVPN
		route. If there is, then an error <bcp14>SHOULD</bcp14> be logged. If
		the v3 flag
		is set (in addition to v2), then the IE flag <bcp14>MUST</bcp14>
		indicate "exclude". If not, then an error <bcp14>SHOULD</bcp14> be lo
		gged. The PE
		<bcp14>MUST</bcp14> generate an IGMP Membership Report for that (*,G)
		and
		each IGMP version in the version flag.
		</li>
		<li>When a PE receives a list of EVPN SMET NLRIs in its BGP update
		message, each with a different source IP address and the same
		Multicast Group Address, and the version flag is set to v3, then the
		PE generates an IGMPv3 Membership Report with a record corresponding
		to the list of source IP addresses and the group address, along with
		the proper indication of inclusion/exclusion.
		</li>
		<li>Upon receiving an EVPN SMET route(s) and before generating the
		corresponding IGMP Membership Request(s), the PE checks to see whethe
		r it has a
		Customer Edge (CE) multicast router for that BD on any of its ESs . T
		he PE provides
		such a check by listening for PIM Hello messages on that AC, i.e.,
		(ES,BD). If the PE does have the router's ACs, then the generated
		IGMP Membership Request(s) is sent to those ACs. If it doesn't have a
		ny of the
		router's ACs, then no IGMP Membership Request(s) needs to be generate
		d. This is
		because sending IGMP Membership Requests to other hosts can result in
		unintentionally preventing a host from joining a specific multicast
		group using IGMPv2, i.e., if the PE does not receive a Membership Rep
		ort from the
		host, it will not forward multicast data to it. Per <xref target="RFC
		4541"
		format="default"/> , when an
		IGMPv2 host receives a Membership Report for a group address that it
		intends to join, the host will suppress its own membership report for
		the same group, and if the PE does not receive an IGMP Membership Rep
		ort from the host,
		it will not forward multicast data to it. In other words, an IGMPv2
		Membership Report <bcp14>MUST NOT</bcp14> be sent on an AC that does
		not lead to a CE
		multicast router. This message suppression is a requirement for IGMPv
		2 hosts.
		This is not a problem for hosts running IGMPv3, because there is no
		suppression of IGMP Membership Reports.
		</li>
		</ol>
		</section>
		<section numbered="true" toc="default">
		<name>IGMP/MLD Leave Group Advertisement in BGP</name>
		<t>When a PE wants to withdraw an EVPN SMET route corresponding to an
		IGMPv2 Leave Group or IGMPv3 "Leave" equivalent message, it
		follows the rules below. The first rule defines the procedure at the
		originator PE, and the last two rules talk about procedures at the remo
		te PE:
		</t>
		<t>Processing at the BGP route originator:
		</t>
		<ol spacing="normal" type="1">
		<li>When a PE receives an IGMPv2 Leave Group or its "Leave" equivalen
		t
		message for IGMPv3 from its attached host, it checks to see if this
		host is the last host that is interested in this multicast group by
		sending a query for the multicast group.


	<t>	<!--[rfced] As "SMET" is expanded as "Selective Multicast Ethernet Tag", may we
	IGMP: Internet Group Management Protocol	remove "Multicast" to avoid redundancy?
	</t>
	<t>
	IR: Ingress Replication
	</t>


	<t>	Original:
	MLD: Multicast Listener Discovery	If the host was indeed the last one (i.e. no responses are
	</t>	received for the query), then the PE MUST re-advertises EVPN
		SMET Multicast route with the corresponding version flag reset.


	<t> OIF: Outgoing Interface for multicast. It can be phys	Perhaps:
	ical interface, virtual interface or tunnel.</t>	If the host was indeed the last one (i.e., no responses are
	<t>	received for the query), then the PE MUST re-advertise the EVPN
	PE: Provider Edge.	SMET route with the corresponding version flag reset.
	</t>	-->


	<t>	If the host was indeed the
	POD: Point of Delivery	last one (i.e., no responses are received for the query), then the PE
	</t>	<bcp14>MUST</bcp14> re-advertise the EVPN SMET Multicast route with t
		he corresponding
		version flag reset. If this is the last version flag to be reset,
		then instead of re-advertising the EVPN route with all version flags
		reset, the PE <bcp14>MUST</bcp14> withdraw the EVPN route for that (*
		,G).
		</li>
		</ol>
		<t>Processing at the BGP route receiver:
		</t>
		<ol spacing="normal" type="1">
		<li>When a PE receives an EVPN SMET route for a given (*,G), it
		compares the received version flags from the route with its per-PE
		stored version flags.


	<t> S-PMSI: Selective P-Multicast Service Interface - a c	<!--[rfced] Is there a word(s) missing before "attached"? Please let us
	onceptual interface for a	know if the perhaps text captures the intended meaning.
	PE to send customer multicast traffic to some of
	the PEs in the same VPN.
	</t>
	<t>
	Single-Active Redundancy Mode: When only a single PE,
	among all the
	PEs attached to an Ethernet segment, is allowed to for
	ward traffic
	to/from that Ethernet segment for a given VLAN, then t
	he Ethernet
	segment is defined to be operating in Single-Active re
	dundancy mode.
	</t>


	<t> SMET: Selective Multicast Eth	Original:
	ernet Tag	If the PE finds that a version flag
	</t>	associated with the (*,G) for the remote PE is reset, then the PE
	<t>	MUST generate IGMP Leave for that (*,G) toward its local
	ToR: Top of Rack	interface (if any) attached to the multicast router for that
	</t>	multicast group.


	</list>	Perhaps:
	</t>	If the PE finds that a version flag
	<t>	associated with the (*,G) for the remote PE is reset, then the PE
	This document also assumes familiarity with the termin	MUST generate IGMP Leave for that (*,G) toward its local
	ology of	interface (if any), which is attached to the multicast router for that
	<xref target="RFC7432"/>, <xref target="RFC3376"/>, <x	multicast group.
	ref target="RFC2236"/> . Though most of the place this document uses term IGMP	-->
	Membership Report, the text applies equally for MLD
	Membership Report too. Similarly, text for IGMPv2 appl
	ies to MLDv1
	and text for IGMPv3 applies to MLDv2. IGMP / MLD vers
	ion encoding in
	BGP update is stated in <xref target="bgp-encoding"/>
	</t>
	<t> It is important to note when there is text considerin
	g whether a PE indicates support for IGMP proxying, the corresponding behavior h
	as a natural analogue for indication of support for MLD proxying, and the analog
	ous requirements apply as well.
	</t>


		If the PE finds that a version flag associated
		with the (*,G) for the remote PE is reset, then the PE <bcp14>MUST</b
		cp14> generate
		IGMP Leave for that (*,G) toward its local interface (if any)
		attached to the multicast router for that multicast group. It should
		be noted that the received EVPN route <bcp14>MUST</bcp14> have at lea
		st one
		version flag set. If all version flags are reset, it is an error
		because the PE should have received an EVPN route withdraw for the
		last version flag. An error <bcp14>MUST</bcp14> be considered as a BG
		P error, and
		the PE <bcp14>MUST</bcp14> apply the
		"treat-as-withdraw" procedure per <xref target="RFC7606" format="defa
		ult"/>.
		</li>
		<li>When a PE receives an EVPN SMET route withdraw, it removes the
		remote PE from its OIF list for that multicast group, and if there ar
		e
		no more OIF entries for that multicast group (either locally or
		remotely), then the PE <bcp14>MUST</bcp14> stop responding to Members
		hip
		Queries from the
		locally attached router (if any). If there is a source for that
		multicast group, the PE stops sending multicast traffic for that sour
		ce.
		</li>
		</ol>
		</section>
	</section>	</section>

		<section numbered="true" toc="default">
	<section title="IGMP/MLD Proxy">	<name>Proxy Querier</name>
	<t>	<t>As mentioned in the previous sections, each PE <bcp14>MUST</bcp14> ha
	The IGMP Proxy mechanism is used to reduce the flooding	ve proxy
	of IGMP	querier functionality for the following reasons:
	messages over an EVPN network similar to ARP proxy used	</t>
	in reducing	<ol spacing="normal" type="1">
	the flooding of ARP messages over EVPN. It also provides	<li>to enable the collection of EVPN PEs providing Layer 2 Virtual Priv
	a triggering	ate Network
	mechanism for the PEs to setup their underlay multicast	(L2VPN) service to
	tunnels. The	act as a distributed multicast router with an anycast IP address for al
	IGMP Proxy mechanism consists of two components:	l
	<list style="numbers">	attached hosts in that subnet</li>
	<t> Proxy for IGMP Membership Reports. </t>	<li>to enable suppression of IGMP Membership Reports and Membership Qu
	<t> Proxy for IGMP Membership Queries. </t>	eries over
	</list>	MPLS/IP core</li>
	</t>	</ol>
	<t>
	The goal of IGMP and MLD proxying is to make th
	e EVPN behave seamlessly for
	the tenant systems with respect to multicast o
	perations, while using a more
	efficient delivery system for signaling and de
	livery across the VPN.
	Accordingly, group state must be tracked synch
	ronously among the PEs
	serving the VPN, with join and leave events pr
	opagated to the peer PEs, and
	each PE tracking the state of each of its peer
	PEs with respect whether
	there are locally attached group members (and
	in some cases, senders), what
	version(s) of IGMP/MLD are in use for those lo
	cally attached group members,
	etc. In order to perform this translation, ea
	ch PE acts as an IGMP router
	for the locally attached domain, and maintains
	the requisite state on
	locally attached nodes, sends periodic members
	hip queries, etc. The role
	of EVPN SMET route propagation is to ensure th
	at each PE's local state is
	propagated to the other PEs so that they share
	a consistent view of the
	overall IGMP Membership Request and Leave Grou
	p state. It is important to
	note that the need to keep such local state ca
	n be triggered by either
	local IGMP traffic or BGP EVPN signaling. In
	most cases a local IGMP event
	will need to be signaled over EVPN, though sta
	te initiated by received EVPN
	traffic will not always need to be relayed to
	the locally attached domain.
	</t>

	<section title="Proxy Reporting">
	<t>
	When IGMP protocol is used between hosts and th
	eir first hop EVPN
	router (EVPN PE), Proxy-reporting is used by th
	e EVPN PE to summarize
	(when possible) reports received from downstrea
	m hosts and propagate
	them in BGP to other PEs that are interested in
	the information. This
	is done by terminating the IGMP Reports in the
	first hop PE, and
	translating and exchanging the relevant informa
	tion among EVPN BGP
	speakers. The information is again translated b
	ack to IGMP message at
	the recipient EVPN speaker. Thus it helps creat
	e an IGMP overlay
	subnet using BGP. In order to facilitate such a
	n overlay, this
	document also defines a new EVPN route type NLR
	I, the EVPN Selective
	Multicast Ethernet Tag route, along with its pr
	ocedures to help
	exchange and register IGMP multicast groups <xr
	ef target="bgp-encoding"/>.

	</t>

	<section title="IGMP/MLD Membership Report Advertisemen
	t in BGP">
	<t>
	When a PE wants to advertise an IGMP
	Membership Report using
	the BGP EVPN route, it follows the fo
	llowing rules (BGP encoding
	stated in <xref target="bgp-encoding"
	/>). Where first four rules are applicable to originator PE and last three rules
	are applicable to remote PE processing SMET routes:

	</t>
	<t>
	Processing at BGP route originator:
	<list style="numbers">
	<t>
	When the first hop PE re
	ceives IGMP Membership Reports
	, belonging to the sam
	e IGMP version, from different attached
	hosts for the same (*,
	G) or (S,G), it SHOULD send a single
	BGP message correspond
	ing to the very first IGMP Membership Request (BGP update as
	soon as possible) for
	that (*,G) or (S,G). This is because BGP is a
	stateful protocol and
	no further transmission of the same report is
	needed. If the IGMP Me
	mbership Request is for (*,G), then multicast group address
	MUST be sent along wit
	h the corresponding version flag (v2 or v3)
	set. In case of IGMPv3
	, the exclude flag MUST also be set to
	indicate that no sourc
	e IP address must be excluded (include all
	sources "*").

	If the IGMP Membership
	Report is for (S,G), then besides setting multicast group
	address along with the
	version flag v3, the source IP address and the
	IE flag MUST be set. I
	t should be noted that when
	advertising the EVPN r
	oute for (S,G), the only valid version flag is
	v3 (v2 flags MUST be s
	et to zero).
	</t>
	<t>

	When the first hop PE
	receives an IGMPv3 Membership Report for (S,G) on a given
	BD, it MUST advertise
	the corresponding EVPN Selective Multicast
	Ethernet Tag (SMET) ro
	ute regardless of whether the source (S) is
	attached to itself or
	not in order to facilitate the source move in
	the future.
	</t>
	<t>

	When the first hop PE
	receives an IGMP version-X Membership Report first for
	(*,G) and then later i
	t receives an IGMP version-Y Membership Report for the same
	(*,G), then it MUST re
	-advertise the same EVPN SMET route with flag
	for version-Y set in a
	ddition to any previously-set version flag(s).
	In other words, the fi
	rst hop PE MUST NOT withdraw the EVPN route
	before sending the new
	route because the flag field is not part of
	BGP route key processi
	ng.

	</t>
	<t>
	When the first hop PE
	receives an IGMP version-X Membership Report first for
	(*,G) and then later i
	t receives an IGMPv3 Membership Report for the same
	multicast group addres
	s but for a specific source address S, then the
	PE MUST advertise a ne
	w EVPN SMET route with v3 flag set (and v2 reset).
	The IE flag also need
	to be set accordingly.
	Since source IP addres
	s is used as part of BGP route key processing
	it is considered as a
	new BGP route advertisement. When different version
	of IGMP Membership Rep
	ort are received, final state MUST be as per section 5.1 of <xref target="RFC337
	6"/>.
	At the end of route pr
	ocessing local and remote group record state MUST be
	as per section 5.1 of
	<xref target="RFC3376"/>.
	</t>
	</list>

	Processing at BGP route r
	eceiver:

	<list style="numbers">
	<t>

	When a PE receives an
	EVPN SMET route with more than one version
	flag set, it will gene
	rate the corresponding IGMP report for (*,G)
	for each version speci
	fied in the flags field. With multiple version
	flags set, there must
	not be source IP address in the received EVPN
	route. If there is, th
	en an error SHOULD be logged. If the v3 flag
	is set (in addition to
	v2), then the IE flag MUST
	indicate "exclude". If
	not, then an error SHOULD be logged. The PE
	MUST generate an IGMP
	Membership Report for that (*,G) and
	each IGMP version in t
	he version flag.
	</t>
	<t>

	When a PE receives a l
	ist of EVPN SMET NLRIs in its BGP update
	message, each with a d
	ifferent source IP address and the same
	multicast group addres
	s, and the version flag is set to v3, then the
	PE generates an IGMPv3
	Membership Report with a record corresponding
	to the list of source
	IP addresses and the group address along with
	the proper indication
	of inclusion/exclusion.
	</t>
	<t>
	Upon receiving EV
	PN SMET route(s) and before generating the
	corresponding IGMP Mem
	bership Request(s), the PE checks to see whether it has any
	CE multicast router fo
	r that BD on any of its ES's . The PE provides
	such a check by listen
	ing for PIM Hello messages on that AC (i.e,
	ES,BD). If the PE does
	have the router's ACs, then the generated
	IGMP Membership Reques
	t(s) are sent to those ACs. If it doesn't have any of the
	router's AC, then no I
	GMP Membership Request(s) needs to be generated. This is
	because sending IGMP M
	embership Requests to other hosts can result in
	unintentionally preven
	ting a host from joining a specific multicast
	group using IGMPv2 - i
	.e., if the PE does not receive a Membership Report from the
	host it will not forwa
	rd multicast data to it. Per <xref target="RFC4541"/> , when an
	IGMPv2 host receives a
	Membership Report for a group address that it
	intends to join, the h
	ost will suppress its own membership report for
	the same group, and if
	the PE does not receive an IGMP Membership Report from the host
	it will not forward mu
	lticast data to it. In other words, an IGMPv2
	Membership Report MUST
	NOT be sent on an AC that does not lead to a CE multicast
	router. This message s
	uppression is a requirement for IGMPv2 hosts.
	This is not a problem
	for hosts running IGMPv3 because there is no
	suppression of IGMP Me
	mbership Reports.

	</t>
	</list>

	</t>
	</section>

	<section title="IGMP/MLD Leave Group Advertisement in B
	GP">
	<t>
	When a PE wants to withdraw an EVPN
	SMET route corresponding to an
	IGMPv2 Leave Group or IGMPv3 "Leave"
	equivalent message, it
	follows the following rules, where f
	irst rule defines the procedure at originator PE and last two rules talk about p
	rocedures at remote PE:
	</t>
	<t>
	Processing at BGP route orig
	inator:

	<list style="numbers">
	<t>
	When a PE receives an
	IGMPv2 Leave Group or its "Leave" equivalent
	message for IGMPv3 fro
	m its attached host, it checks to see if this
	host is the last host
	that is interested in this multicast group by
	sending a query for th
	e multicast group. If the host was indeed the
	last one (i.e. no resp
	onses are received for the query), then the PE
	MUST re-advertises EVP
	N SMET Multicast route with the corresponding
	version flag reset. If
	this is the last version flag to be reset,
	then instead of re-adv
	ertising the EVPN route with all version flags
	reset, the PE MUST wit
	hdraw the EVPN route for that (*,G).
	</t>
	</list>
	Processing at BGP route r
	eceiver:
	<list style="numbers">
	<t>
	When a PE receives an
	EVPN SMET route for a given (*,G), it
	compares the received
	version flags from the route with its per-PE
	stored version flags.
	If the PE finds that a version flag associated
	with the (*,G) for the
	remote PE is reset, then the PE MUST generate
	IGMP Leave for that (*
	,G) toward its local interface (if any)
	attached to the multic
	ast router for that multicast group. It should
	be noted that the rece
	ived EVPN route MUST at least have one
	version flag set. If a
	ll version flags are reset, it is an error
	because the PE should
	have received an EVPN route withdraw for the
	last version flag. Err
	or MUST be considered as a BGP error and the PE MUST apply the
	"treat-as-withdraw" pr
	ocedure of <xref target="RFC7606"/>.
	</t>
	<t>
	When a PE receives an
	EVPN SMET route withdraw, it removes the
	remote PE from its OIF
	list for that multicast group and if there are
	no more OIF entries fo
	r that multicast group (either locally or
	remotely), then the PE
	MUST stop responding to Membership Queries from the
	locally attached route
	r (if any). If there is a source for that
	multicast group, the P
	E stops sending multicast traffic for that
	source.
	</t>
	</list>
	</t>
	</section>
	</section>

	<section title="Proxy Querier">
	<t>
	As mentioned in the previous sections, each PE
	MUST have proxy
	querier functionality for the following reasons
	:

	<list style="numbers">
	<t>
	To enable the collection of EVP
	N PEs providing L2VPN service to
	act as distributed multicast ro
	uter with Anycast IP address for all
	attached hosts in that subnet.
	</t>

	<t>
	To enable suppression of IGMP M
	embership Reports and Membership Queries over
	MPLS/IP core.
	</t>
	</list>
	</t>
	</section>

	</section>	</section>

		</section>
	<section title="Operation">	<section numbered="true" toc="default">
	<t>	<name>Operation</name>
	Consider the EVPN network of Figure-1, where there is an EV	<t>Consider the EVPN network in <xref target="EVPN"/>, where there is an E
	PN	VPN
	instance configured across the PEs shown in this figur	instance configured across the PEs (namely PE1,
	e (namely PE1,	PE2, and PE3). Let's consider that this EVPN instance consists of a
	PE2, and PE3). Let's consider that this EVPN instance	single bridge domain (single subnet) with all the hosts and sources and
	consists of a	the multicast router connected to this subnet. PE1 only has hosts (host de
	single bridge domain (single subnet) with all the host	noted by Hx)
	s, sources, and	connected to it. PE2 has a mix of hosts and a multicast source. PE3
	the multicast router connected to this subnet. PE1 onl	has a mix of hosts, a multicast source (source denoted by Sx), and a multi
	y has hosts(host denoted by Hx)	cast router
	connected to it. PE2 has a mix of hosts and a multicas	(router denoted by Rx).
	t source. PE3	Furthermore, let's consider that for (S1,G1), R1 is used as the
	has a mix of hosts, a multicast source (source denoted	multicast router. The following subsections describe the IGMP proxy
	by Sx), and a multicast router (router denoted by Rx).	operation in different PEs with regard to whether the locally
	Furthermore, let's consider that for (S1,G1), R1 is us	attached devices for that subnet are:
	ed as the	</t>
	multicast router. The following subsections describe t	<ul spacing="normal">
	he IGMP proxy	<li>only hosts,</li>
	operation in different PEs with regard to whether the	<li>a mix of hosts and a multicast source, or</li>
	locally	<li>a mix of hosts, a multicast source, and a multicast router.</li>
	attached devices for that subnet are:	</ul>
		<figure anchor="EVPN">
	<list style="symbols">	<name>EVPN Network</name>
	<t>	<artwork name=">EVPN network" type="" align="center" alt=""><![CDATA[
	only hosts	+--------------+
	</t>	\| \|
	<t>	\| \|
	mix of hosts and multicast source	+----+ \| \| +----+
	</t>	H1:(,G1)v2 ---\| \| \| \| \| \|---- H6(,G1)v2
	<t>	H2:(*,G1)v2 ---\| PE1\| \| IP/MPLS \| \| PE2\|---- H7(S2,G2)v3
	mix of hosts, multicast source, and mu	H3:(*,G1)v3 ---\| \| \| Network \| \| \|---- S2
	lticast router	H4:(S2,G2)v3 --\| \| \| \| \| \|
	</t>	+----+ \| \| +----+
	</list>	\| \|
	</t>	+----+ \| \|
		H5:(S1,G1)v3 --\| \| \| \|
	<figure >	S1 ---\| PE3\| \| \|
	<artwork ><![CDATA[	R1 ---\| \| \| \|
		+----+ \| \|
	+--------------+	\| \|
	\| \|	+--------------+
	\| \|	]]></artwork>
	+----+ \| \| +----+
	H1:(,G1)v2 ---\| \| \| \| \| \|---- H6(,G1)v2
	H2:(*,G1)v2 ---\| PE1\| \| IP/MPLS \| \| PE2\|---- H7(S2,G2)v3
	H3:(*,G1)v3 ---\| \| \| Network \| \| \|---- S2
	H4:(S2,G2)v3 --\| \| \| \| \| \|
	+----+ \| \| +----+
	\| \|
	+----+ \| \|
	H5:(S1,G1)v3 --\| \| \| \|
	S1 ---\| PE3\| \| \|
	R1 ---\| \| \| \|
	+----+ \| \|
	\| \|
	+--------------+

	Figure 1: EVPN network

	]]></artwork>
	</figure>

	<section title="PE with only attached hosts for a given subnet">
	<t>
	When PE1 receives an IGMPv2 Membership Report from H1, it
	does not forward
	this Membership Report to any of its other ports (for thi
	s subnet) because all
	these local ports are associated with the hosts. PE1 send
	s an
	EVPN Multicast Group route corresponding to this Membersh
	ip Report for (*,G1) and
	setting v2 flag. This EVPN route is received by PE2 and P
	E3 that are
	the members of the same BD (i.e., same EVI in case of VLA
	N-based
	service or EVI,VLAN in case of VLAN-aware bundle service)
	. PE3
	reconstructs the IGMPv2 Membership Report from this EVPN
	BGP route and only
	sends it to the port(s) with multicast routers attached t
	o it (for
	that subnet). In this example, PE3 sends the reconstructe
	d IGMPv2
	Membership Report for (*,G1) only to R1. Furthermore, ev
	en though PE2
	receives the EVPN BGP route, it does not send it to any o
	f its ports
	for that subnet; viz, ports associated with H6 and H7.
	</t>

	<t>
	When PE1 receives the second IGMPv2 Membership Report fro
	m H2 for the same
	multicast group (*,G1), it only adds that port to its OIF
	list but it
	doesn't send any EVPN BGP route because there is no chang
	e in
	information. However, when it receives the IGMPv3 Members
	hip Report from H3 for
	the same (*,G1). Besides adding the corresponding port to
	its OIF
	list, it re-advertises the previously sent EVPN SMET rout
	e with the
	v3 and exclude flag set.
	</t>

	<t>
	Finally when PE1 receives the IGMPv3 Membership Report fr
	om H4 for (S2,G2), it
	advertises a new EVPN SMET route corresponding to it.
	</t>
	</section>

	<section title="PE with a mix of attached hosts and multicast source">
	<t>
	The main difference in this case is that when PE2 rece
	ives the IGMPv3
	Membership Report from H7 for (S2,G2), it does adverti
	se it in BGP to support
	source move even though PE2 knows that S2 is attached
	to its local
	AC. PE2 adds the port associated with H7 to its OIF li
	st for (S2,G2).
	The processing for IGMPv2 received from H6 is the same
	as the IGMPv2
	Membership Report described in previous section.
	</t>

	</section>

	<section title="PE with a mix of attached hosts, a multicast source and a
	router">
	<t>
	The main difference in this case relative to the previ
	ous two
	sections is that IGMP v2/v3 Membership Report messages
	received locally need to
	be sent to the port associated with router R1. Further
	more, the Membership Reports
	received via BGP (SMET) need to be passed to the R1 po
	rt but filtered
	for all other ports.
	</t>
	</section>
	</section>

	<section title="All-Active Multi-Homing">
	<t>
	Because the LAG flow hashing algorithm used by the CE is unkno
	wn at
	the PE, in an All-Active redundancy mode it must be assumed th
	at the
	CE can send a given IGMP message to any one of the multi-homed
	PEs,
	either DF or non-DF; i.e., different IGMP Membership Request m
	essages can arrive at
	different PEs in the redundancy group and furthermore their
	corresponding Leave messages can arrive at PEs that are differ
	ent
	from the ones that received the Membership Report. Therefore,
	all PEs
	attached to a given ES must coordinate IGMP Membership Request
	and Leave Group
	(x,G) state, where x may be either '*' or a particular source
	S, for
	each BD on that ES. Each PE has a local copy of that state and
	the EVPN signaling serves to synchronize state across PEs. This allows the DF f
	or that (ES,BD) to correctly
	advertise or withdraw a Selective Multicast Ethernet Tag (SMET
	) route
	for that (x,G) group in that BD when needed.
	All-Active multihoming PEs for a given ES MUST support IGMP
	synchronization procedures described in this section if they n
	eed to
	perform IGMP proxy for hosts connected to that ES.
	</t>

	<section title="Local IGMP/MLD Membership Report Synchronization">
	<t>
	When a PE, either DF or non-DF, receives on a given
	multihomed ES
	operating in All-Active redundancy mode, an IGMP Me
	mbership Report
	for (x,G), it determines the BD to which the IGMP M
	embership Report
	belongs. If the PE doesn't already have local IGMP
	Membership Request (x,G) state
	for that BD on that ES, it MUST instantiate local I
	GMP Membership Request (x,G)
	state and MUST advertise a BGP IGMP Membership Repo
	rt Synch route for that (ES,BD).
	Local IGMP Membership Request (x,G) state refers t
	o IGMP Membership Request (x,G) state
	that is created as a result of processing an IGMP
	Membership Report
	for (x,G).
	</t>

	<t>
	The IGMP Membership Report Synch route MUST carry the
	ES-Import RT for the ES on
	which the IGMP Membership Report was received. Thus i
	t MUST only be
	imported by the PEs attached to that ES and not any ot
	her PEs.
	</t>

	<t>
	When a PE, either DF or non-DF, receives an IGMP Me
	mbership Report Synch route it
	installs that route and if it doesn't already have
	IGMP Membership Request (x,G)
	state for that (ES,BD), it MUST instantiate that IG
	MP Membership Request (x,G)
	state - i.e., IGMP Membership Request (x,G) state i
	s the union of the local IGMP
	Membership Report (x,G) state and the installed IGM
	P Membership Report Synch route. If the DF
	did not already advertise (originate) a SMET route
	for that (x,G)
	group in that BD, it MUST do so now.
	</t>

	<t>
	When a PE, either DF or non-DF, deletes its local IGM
	P Membership Request (x,G)
	state for that (ES,BD), it MUST withdraw its BGP IGMP
	Membership Report Synch
	route for that (ES,BD).
	</t>

	<t>
	When a PE, either DF or non-DF, receives the withdr
	awal of an IGMP
	Membership Report Synch route from another PE it MU
	ST remove that route. When a
	PE has no local IGMP Membership Request (x,G) state
	and it has no installed IGMP
	Membership Report Synch routes, it MUST remove IGMP
	Membership Request (x,G) state for that (ES,BD).
	If the DF no longer has IGMP Membership Request (x,
	G) state for that BD on
	any ES for which it is DF, it MUST withdraw its SME
	T route for that
	(x,G) group in that BD.
	</t>
	<t>
	In other words, a PE advertises an SMET route for t
	hat (x,G) group in
	that BD when it has IGMP Membership Request (x,G) s
	tate in that BD on at least one
	ES for which it is DF and it withdraws that SMET ro
	ute when it does
	not have IGMP Membership Request (x,G) state in tha
	t BD on any ES for which it is
	DF.
	</t>
	</section>

	<section title="Local IGMP/MLD Leave Group Synchronization">
	<t>
	When a PE, either DF or non-DF, receives, on a given
	multihomed ES
	operating in All-Active redundancy mode, an IGMP Lea
	ve Group message
	for (x,G) from the attached CE, it determines the BD
	to which the
	IGMPv2 Leave Group belongs. Regardless of whether i
	t has IGMP Membership Request
	(x,G) state for that (ES,BD), it initiates the (x,G)
	leave group
	synchronization procedure, which consists of the fol
	lowing steps:

	<list style="numbers">
	<t>
	It computes the Maximum Response Tim
	e, which is the duration of
	(x,G) leave group synchronization pr
	ocedure. This is the product of
	two locally configured values, Last
	Member Query Count and Last
	Member Query Interval (described in
	Section 3 of <xref target="RFC2236"/>), plus a
	delta corresponding to the time it t
	akes for a BGP advertisement to
	propagate between the PEs attached t
	o the multihomed ES (delta is a
	consistently configured value on all
	PEs attached to the multihomed
	ES).
	</t>

	<t>
	It starts the Maximum Response T
	ime timer. Note that the receipt
	of subsequent IGMP Leave Group m
	essages or BGP Leave Synch routes for
	(x,G) do not change the value of
	a currently running Maximum Response
	Time timer and are ignored by th
	e PE.
	</t>

	<t>
	It initiates the Last Member Que
	ry procedure described in Section
	3 of <xref target="RFC2236"/>; v
	iz, it sends a number of Group-Specific Query (x,G)
	messages (Last Member Query Coun
	t) at a fixed interval (Last Member
	Query Interval) to the attached
	CE.
	</t>

	<t>
	It advertises an IGMP Leave Sync
	h route for that that (ES,BD).
	This route notifies the other mu
	ltihomed PEs attached to the given
	multihomed ES that it has initia
	ted an (x,G) leave group
	synchronization procedure; i.e.,
	it carries the ES-Import RT for the
	ES on which the IGMP Leave Group
	was received. It also contains the
	Maximum Response Time.
	</t>
	<t>
	When the Maximum Response Timer
	expires, the PE that has
	advertised the IGMP Leave Synch
	route withdraws it.
	</t>
	</list>
	</t>

	<section title="Remote Leave Group Synchronization">
	<t>
	When a PE, either DF or non-DF, receives an IG
	MP Leave Synch route it
	installs that route and it starts a timer for
	(x,G) on the specified
	(ES,BD) whose value is set to the Maximum Resp
	onse Time in the
	received IGMP Leave Synch route. Note that th
	e receipt of subsequent
	IGMPv2 Leave Group messages or BGP Leave Synch
	routes for (x,G) do
	not change the value of a currently running Ma
	ximum Response Time
	timer and are ignored by the PE.
	</t>
	</section>

	<section title="Common Leave Group Synchronization">
	<t>
	If a PE attached to the multihomed ES receives
	an IGMP Membership
	Report for (x,G) before the Maximum Response Ti
	me timer expires, it
	advertises a BGP IGMP Membership Report Synch r
	oute for that (ES,BD). If it
	doesn't already have local IGMP Membership Requ
	est (x,G) state for that (ES,BD),
	it instantiates local IGMP Membership Request (
	x,G) state. If the DF is not
	currently advertising (originating) a SMET rout
	e for that (x,G) group
	in that BD, it does so now.
	</t>

	<t>
	If a PE attached to the multihomed ES receiv
	es an IGMP Membership Report Synch
	route for (x,G) before the Maximum Response
	Time timer expires, it
	installs that route and if it doesn't alread
	y have IGMP Membership Request (x,G)
	state for that BD on that ES, it instantiate
	s that IGMP Membership Request (x,G)
	state. If the DF has not already advertised
	(originated) a SMET route
	for that (x,G) group in that BD, it does so
	now.
	</t>

	<t>
	When the Maximum Response Timer expires a PE
	that has advertised an
	IGMP Leave Synch route, withdraws it. Any P
	E attached to the
	multihomed ES, that started the Maximum Resp
	onse Time and has no
	local IGMP Membership Request (x,G) state an
	d no installed IGMP Membership Report Synch routes,
	it removes IGMP Membership Request (x,G) sta
	te for that (ES,BD). If the DF no
	longer has IGMP Membership Request (x,G) sta
	te for that BD on any ES for which it
	is DF, it withdraws its SMET route for that
	(x,G) group in that BD.
	</t>
	</section>

	</section>

	<section title="Mass Withdraw of Multicast Membership Report Sync rout
	e in case of failure">
	<t>
	A PE which has received an IGMP Membership Request
	would have synced the IGMP Membership Report
	by the procedure defined in section 6.1. If a PE wi
	th local Membership Report
	state goes down or the PE to CE link goes down, it
	would lead to a
	mass withdraw of multicast routes. Remote PEs (PEs
	where these routes
	were remote IGMP Membership Reports) SHOULD NOT rem
	ove the state immediately;
	instead General Query SHOULD be generated to refres
	h the states.
	There are several ways to detect failure at a
	peer, e.g. using IGP next hop tracking or ES route
	withdraw.
	</t>
	</section>

	</section>

	<section title="Single-Active Multi-Homing">
	<t>
	Note that to facilitate state synchronization after failove
	r, the PEs
	attached to a multihomed ES operating in Single-Active redu
	ndancy mode
	SHOULD also coordinate IGMP Membership Report (x,G) state.
	In this case all IGMP
	Membership Report messages are received by the DF and distr
	ibuted to the non-DF
	PEs using the procedures described above.
	</t>
	</section>

	<section title="Selective Multicast Procedures for IR tunnels">
	<t>
	If an ingress PE uses ingress replication, then for a
	given (x,G)
	group in a given BD:

	<list style="numbers">
	<t>
	It sends (x,G) traffic to the set of P
	Es not supporting IGMP or MLD
	Proxy. This set consists of any PE tha
	t has advertised an IMET route for the BD
	without a Multicast Flags extended co
	mmunity or with a Multicast Flags extended
	community in which neither the IGMP Pr
	oxy support nor the MLD Proxy support flags are set.
	</t>

	<t>
	It sends (x,G) traffic to the set of P
	Es supporting IGMP or MLD Proxy
	and having listeners for that (x,G) gr
	oup in that BD. This set consists of any PE that has advertised an IMET route fo
	r the BD
	with a Multicast Flags extended commun
	ity in which the IGMP Proxy support and/or
	the MLD Proxy support flags are set a
	nd that has advertised a SMET route for that (x,G)
	group in that BD.
	</t>

	</list>

	</t>

	</section>

	<section title="BGP Encoding" anchor="bgp-encoding">
	<t>
	This document defines three new BGP EVPN routes to carry
	IGMP
	Membership Reports. The route types are known as:
	</t>

	<t> + 6 - Selective Multicast Ethernet Tag
	Route </t>
	<t> + 7 - Multicast Membership Report Synch
	Route </t>
	<t> + 8 - Multicast Leave Synch Route </t>

	<t>

	The detailed encoding and procedures for these route t
	ypes are
	described in subsequent sections.
	</t>

	<section title="Selective Multicast Ethernet Tag Route" anchor="
	SMET">
	<t>
	A Selective Multicast Ethernet Tag route type sp
	ecific EVPN NLRI
	consists of the following:
	</t>

	<figure >
	<artwork ><![CDATA[

	+---------------------------------------+
	\| RD (8 octets) \|
	+---------------------------------------+
	\| Ethernet Tag ID (4 octets) \|
	+---------------------------------------+
	\| Multicast Source Length (1 octet) \|
	+---------------------------------------+
	\| Multicast Source Address (variable) \|
	+---------------------------------------+
	\| Multicast Group Length (1 octet) \|
	+---------------------------------------+
	\| Multicast Group Address (Variable) \|
	+---------------------------------------+
	\| Originator Router Length (1 octet) \|
	+---------------------------------------+
	\| Originator Router Address (variable) \|
	+---------------------------------------+
	\| Flags (1 octet) \|
	+---------------------------------------+

	]]></artwork>
	</figure>

	<t>
	For the purpose of BGP route key processing, all the fields ar
	e
	considered to be part of the prefix in the NLRI except for the
	one-
	octet flag field. The Flags fields are defined as follows:
	</t>

	<figure >
	<artwork ><![CDATA[
	0 1 2 3 4 5 6 7
	+--+--+--+--+--+--+--+--+
	\| reserved \|IE\|v3\|v2\|v1\|
	+--+--+--+--+--+--+--+--+

	]]></artwork>
	</figure>	</figure>

		<section numbered="true" toc="default">
		<name>PE with Only Attached Hosts for a Given Subnet</name>
		<t>When PE1 receives an IGMPv2 Membership Report from H1, it does not fo
		rward
		this Membership Report to any of its other ports (for this subnet) becaus
		e all
		these local ports are associated with the hosts.


	<t>	<!-- [rfced] Does PE1 send an EVPN Multicast Group and set the v2 flag
	<list style="symbols">	(Option A), or does the EVPN Multicast Group correspond to the
	<t>	Membership Report and set the v2 flag (Option B)?
	The least significant bit, bit 7 indicates sup
	port for IGMP version
	1. Since IGMP V1 is being deprecated sender MU
	ST set it as 0 for IGMP and
	receiver MUST ignore it.
	</t>
	<t>
	The second least significant bit, bit 6 indica
	tes support for IGMP
	version 2.
	</t>
	<t>
	The third least significant bit, bit 5 indicat
	es support for IGMP
	version 3.
	</t>
	<t>
	The fourth least significant bit, bit 4 indica
	tes whether the (S,G)
	information carried within the route-type is o
	f an Include Group type
	(bit value 0) or an Exclude Group type (bit va
	lue 1). The Exclude
	Group type bit MUST be ignored if bit 5 is not
	set.
	</t>
	<t>
	This EVPN route type is used to carry tenant I
	GMP multicast group
	information. The flag field assists in distrib
	uting IGMP Membership
	Report of a given host for a given multicast r
	oute. The version
	bits help associate IGMP version of receivers
	participating within
	the EVPN domain.
	</t>
	<t>
	The include/exclude (IE) bit helps in creating
	filters for a given
	multicast route.
	</t>
	<t>
	If route is used for IPv6 (MLD) then bit 7 ind
	icates support for MLD
	version 1. The second least significant bit, b
	it 6 indicates support
	for MLD version 2. Since there is no MLD versi
	on 3, in case of IPv6
	route third least significant bit MUST be 0. I
	n case of IPv6 routes,
	the fourth least significant bit MUST be ignor
	ed if bit 6 is not
	set.

	</t>
	<t> Reserved bits MUST be set to 0 by sender. And
	receiver MUST ignore the Reserved bits.
	</t>
	</list>
	</t>

	<section title="Constructing the Selective Multicast Ethernet Tag ro
	ute">
	<t>
	This section describes the procedures used to constr
	uct the Selective
	Multicast Ethernet Tag (SMET) route.
	</t>

	<t>
	The Route Distinguisher (RD) SHOULD be a Type
	1 RD <xref target="RFC4364"/>. The
	value field comprises an IP address of the PE
	(typically, the
	loopback address) followed by a number unique
	to the PE.
	</t>

	<t>
	The Ethernet Tag ID MUST be set as procedure defined
	in <xref target="RFC7432"/>.
	</t>

	<t>
	The Multicast Source Length MUST be set to le
	ngth of the multicast
	Source address in bits. If the Multicast Sour
	ce Address field
	contains an IPv4 address, then the value of t
	he Multicast Source
	Length field is 32. If the Multicast Source A
	ddress field contains an
	IPv6 address, then the value of the Multicast
	Source Length field is
	128. In case of a (*,G) Membership Report, th
	e Multicast Source Length is set to
	0.
	</t>

	<t>
	The Multicast Source Address is the source IP
	address from the IGMP
	Membership Report. In case of a (*,G), this f
	ield is not used.
	</t>

	<t>
	The Multicast Group Length MUST be set to length
	of multicast group
	address in bits. If the Multicast Group Address
	field contains an
	IPv4 address, then the value of the Multicast Gr
	oup Length field is
	32. If the Multicast Group Address field contai
	ns an IPv6 address,
	then the value of the Multicast Group Length fie
	ld is 128.
	</t>

	<t>
	The Multicast Group Address is the Group address
	from the IGMP or MLD
	Membership Report.
	</t>

	<t>
	The Originator Router Length is the length of th
	e Originator Router
	Address in bits.
	</t>

	<t>
	The Originator Router Address is the IP addre
	ss of router originating this route.
	The SMET Originator Router IP address MUST ma
	tch that of the IMET (or S-PMSI AD)
	route originated for the same EVI by the same
	downstream PE.
	</t>

	<t>
	The Flags field indicates the version of IGMP
	protocol from which the
	Membership Report was received. It also indic
	ates whether the
	multicast group had the INCLUDE or EXCLUDE bi
	t set.
	</t>
	<t> Reserved bits MUST be set to 0. They can be defined
	in future by other document.
	</t>


	<t>	Original:
	IGMP is used to receive group membership info	PE1 sends an EVPN Multicast Group route corresponding to this
	rmation from hosts	Membership Report for (*,G1) and setting the v2 flag.
	by TORs. Upon receiving the hosts expression
	of interest of a
	particular group membership, this information
	is then forwarded using
	SMET route. The NLRI also keeps track
	of receiver's IGMP protocol version and any s
	ource filtering for a
	given group membership. All EVPN SMET routes
	are announced with per-
	EVI Route Target extended communities.
	</t>


	</section>	Perhaps:


	<section title="Reconstructing IGMP / MLD Membership Reports fro	A) PE1 sends an EVPN Multicast Group route corresponding to this
	m Selective Multicast Route">	Membership Report for (*,G1) and sets the v2 flag.
	<t> Th
	is section describes the procedures used to reconstruct IGMP / MLD Membership Re
	ports from SMET route.
	</t>
	<t>
	<list style="symbols">
	<t> If multicast group length is 32, rou
	te would be translated to IGMP membership request. If multicast group length is
	128, route would be translated to MLD membership request.
	</t>


	<t> Multicast group address field would be translated	or
	to IGMP / MLD group address.
	</t>
	<t> If Multicast source length is set to zero it would
	be translated to any source (*).
	If multicast source length is non zero, Multica
	st source address field would be translated to IGMP / MLD source address.
	</t>
	<t> If flag bit 7 is set, it translates Membership repo
	rt to be IGMP V1 or MLD V1.
	</t>
	<t> If flag bit 6 is set, it translates Membership repo
	rt to be IGMP V2 or MLD V2.
	</t>
	<t> Flag bit 5 is only valid for IGMP Membership report
	and if it is set, it translates to IGMP V3 report.
	</t>
	<t> If IE flag is set, it translate to IGMP / MLD Exc
	lude mode membership report. If IE flag is not set (zero), it translates to Incl
	ude mode membership report.
	</t>
	</list>
	</t>


	</section>	B) PE1 sends an EVPN Multicast Group route that corresponds to this
		Membership Report for (*,G1) and sets the v2 flag.
		-->
		PE1 sends an
		EVPN Multicast Group route corresponding to this Membership Report for (*
		,G1) and
		setting the v2 flag. This EVPN route is received by PE2 and PE3, which ar
		e
		the members of the same BD (i.e., same EVI in case of a VLAN-based
		service or EVI and VLAN in case of a VLAN-aware bundle service). PE3
		reconstructs the IGMPv2 Membership Report from this EVPN BGP route and on
		ly
		sends it to the port(s) with multicast routers attached to it (for
		that subnet). In this example, PE3 sends the reconstructed IGMPv2
		Membership Report for (*,G1) only to R1. Furthermore, even though PE2
		receives the EVPN BGP route, it does not send it to any of its ports
		for that subnet (viz., ports associated with H6 and H7).
		</t>
		<t>When PE1 receives the second IGMPv2 Membership Report from H2 for the
		same
		multicast group (*,G1), it only adds that port to its OIF list, but it
		doesn't send any EVPN BGP routes because there is no change in
		information. However, when it receives the IGMPv3 Membership Report from
		H3 for
		the same (*,G1), besides adding the corresponding port to its OIF
		list, it re-advertises the previously sent EVPN SMET route with the
		v3 and exclude flag set.
		</t>
		<t>Finally, when PE1 receives the IGMPv3 Membership Report from H4 for (
		S2,G2), it
		advertises a new EVPN SMET route corresponding to it.
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>PE with a Mix of Attached Hosts and a Multicast Source</name>
		<t>The main difference in this case is that when PE2 receives the IGMPv3
		Membership Report from H7 for (S2,G2), it advertises it in BGP to support
		the
		source moving, even though PE2 knows that S2 is attached to its local
		AC. PE2 adds the port associated with H7 to its OIF list for (S2,G2).
		The processing for IGMPv2 received from H6 is the same as the IGMPv2
		Membership Report described in the previous section.
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>PE with a Mix of Attached Hosts, a Multicast Source, and a Router<
		/name>
		<t>The main difference in this case relative to the previous two
		sections is that IGMPv2/v3 Membership Report messages received locally ne
		ed to
		be sent to the port associated with router R1. Furthermore, the Membershi
		p Reports
		received via BGP (SMET) need to be passed to the R1 port but filtered
		for all other ports.
		</t>
		</section>
		</section>
		<section numbered="true" toc="default">
		<name>All-Active Multihoming</name>
		<t>Because the Link Aggregation Group (LAG) flow hashing algorithm used by
		the CE is unknown at
		the PE, in an All-Active redundancy mode, it must be assumed that the
		CE can send a given IGMP message to any one of the multihomed PEs,
		either Designated Forwarder (DF) or non-DF, i.e., different IGMP Membershi
		p
		Request messages can arrive at
		different PEs in the redundancy group. Furthermore, their
		corresponding Leave messages can arrive at PEs that are different
		from the ones that received the Membership Report. Therefore, all PEs
		attached to a given Ethernet Segment (ES) must coordinate the IGMP Members
		hip Request and Leave Group
		(x,G) state, where x may be either "*" or a particular source S for
		each BD on that ES. Each PE has a local copy of that state, and the EVPN s
		ignaling
		serves to synchronize that state across PEs. This allows the DF for that (
		ES,BD) to correctly
		advertise or withdraw a SMET route
		for that (x,G) group in that BD when needed.
		All-Active multihoming PEs for a given ES <bcp14>MUST</bcp14> support IGMP
		synchronization procedures described in this section if they need to
		perform IGMP proxy for hosts connected to that ES.
		</t>
		<section numbered="true" toc="default" anchor="local-igmp-mld">
		<name>Local IGMP/MLD Membership Report Synchronization</name>
		<t>When a PE, either DF or non-DF, receives an IGMP Membership Report
		for (x,G) on a given multihomed ES operating in All-Active redundancy mod
		e, it determines the BD to which the IGMP Membership Report
		belongs. If the PE doesn't already have the local IGMP Membership Request
		(x,G) state
		for that BD on that ES, it <bcp14>MUST</bcp14> instantiate that local IGM
		P Membership
		Request (x,G)
		state and <bcp14>MUST</bcp14> advertise a BGP IGMP Membership Report Sync
		h route
		for that (ES,BD).
		The local IGMP Membership Request (x,G) state refers to the IGMP Membersh
		ip Request (x,G) state
		that is created as a result of processing an IGMP Membership Report
		for (x,G).
		</t>
		<t>The IGMP Membership Report Synch route <bcp14>MUST</bcp14> carry the
		ES-Import
		Route Target (RT) for the ES on
		which the IGMP Membership Report was received. Thus, it <bcp14>MUST</bcp
		14> only be
		imported by the PEs attached to that ES and not any other PEs.
		</t>
		<t>When a PE, either DF or non-DF, receives an IGMP Membership Report Sy
		nch route, it
		installs that route, and if it doesn't already have the IGMP Membership R
		equest (x,G)
		state for that (ES,BD), it <bcp14>MUST</bcp14> instantiate that IGMP Memb
		ership
		Request (x,G)
		state, i.e., the IGMP Membership Request (x,G) state is the union of the
		local IGMP
		Membership Report (x,G) state and the installed IGMP Membership Report Sy
		nch route.
		If the DF did not already advertise (originate) a SMET route for that (x,
		G)
		group in that BD, it <bcp14>MUST</bcp14> do so now.
		</t>
		<t>When a PE, either DF or non-DF, deletes its local IGMP Membership Req
		uest (x,G)
		state for that (ES,BD), it <bcp14>MUST</bcp14> withdraw its BGP IGMP Memb
		ership
		Report Synch route for that (ES,BD).
		</t>
		<t>When a PE, either DF or non-DF, receives the withdrawal of an IGMP
		Membership Report Synch route from another PE, it <bcp14>MUST</bcp14> rem
		ove that route.
		When a PE has no local IGMP Membership Request (x,G) state and it has no
		installed IGMP
		Membership Report Synch routes, it <bcp14>MUST</bcp14> remove that IGMP M
		embership Request
		(x,G) state for that (ES,BD).
		If the DF no longer has the IGMP Membership Request (x,G) state for that
		BD on
		any ES for which it is the DF, it <bcp14>MUST</bcp14> withdraw its SMET r
		oute for that
		(x,G) group in that BD.
		</t>
		<t>In other words, a PE advertises a SMET route for that (x,G) group in
		that BD when it has the IGMP Membership Request (x,G) state on at least o
		ne
		ES for which it is the DF, and it withdraws that SMET route when it does
		not have an IGMP Membership Request (x,G) state in that BD on any ES for
		which it is
		the DF.
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>Local IGMP/MLD Leave Group Synchronization</name>
		<t>When a PE, either DF or non-DF, receives an IGMP Leave Group message
		for (x,G) from the attached CE on a given multihomed ES
		operating in All-Active redundancy mode, it determines the BD to which th
		e
		IGMPv2 Leave Group belongs. Regardless of whether it has the IGMP Member
		ship Request
		(x,G) state for that (ES,BD), it initiates the (x,G) leave group
		synchronization procedure, which consists of the following steps:
		</t>
		<ol spacing="normal" type="1">
		<li>It computes the Maximum Response Time, which is the duration of the
		(x,G) leave group synchronization procedure. This is the product of
		two locally configured values, Last Member Query Count and Last
		Member Query Interval (described in <xref target="RFC2236" section="3"
		sectionFormat="of"/>), plus a
		delta corresponding to the time it takes for a BGP advertisement to
		propagate between the PEs attached to the multihomed ES (delta is a
		consistently configured value on all PEs attached to the multihomed
		ES).
		</li>
		<li>It starts the Maximum Response Time timer. Note that the receipt
		of subsequent IGMP Leave Group messages or BGP Leave Synch routes for
		(x,G) do not change the value of a currently running Maximum Response
		Time timer and are ignored by the PE.
		</li>
		<li>It initiates the Last Member Query procedure described in
		<xref target="RFC2236" section="3" sectionFormat="of"/>; viz., it
		sends a number of Group-Specific Query (x,G)
		messages (Last Member Query Count) at a fixed interval (Last Member
		Query Interval) to the attached CE.</li>
		<li>It advertises an IGMP Leave Synch route for that (ES,BD).
		This route notifies the other multihomed PEs attached to the given
		multihomed ES that it has initiated an (x,G) leave group
		synchronization procedure, i.e., it carries the ES-Import RT for the
		ES on which the IGMP Leave Group was received. It also contains the
		Maximum Response Time.
		</li>
		<li>When the Maximum Response Time timer expires, the PE that has
		advertised the IGMP Leave Synch route withdraws it.
		</li>
		</ol>
		<section numbered="true" toc="default">
		<name>Remote Leave Group Synchronization</name>
		<t>When a PE, either DF or non-DF, receives an IGMP Leave Synch route,
		it
		installs that route and it starts a timer for (x,G) on the specified
		(ES,BD), whose value is set to the Maximum Response Time in the
		received IGMP Leave Synch route. Note that the receipt of subsequent
		IGMPv2 Leave Group messages or BGP Leave Synch routes for (x,G) do
		not change the value of a currently running Maximum Response Time
		timer and are ignored by the PE.
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>Common Leave Group Synchronization</name>
		<t>If a PE attached to the multihomed ES receives an IGMP Membership
		Report for (x,G) before the Maximum Response Time timer expires, it
		advertises a BGP IGMP Membership Report Synch route for that (ES,BD). I
		f it
		doesn't already have the local IGMP Membership Request (x,G) state for
		that (ES,BD),
		it instantiates that local IGMP Membership Request (x,G) state. If the
		DF is not
		currently advertising (originating) a SMET route for that (x,G) group
		in that BD, it does so now.
		</t>
		<t>If a PE attached to the multihomed ES receives an IGMP Membership R
		eport Synch
		route for (x,G) before the Maximum Response Time timer expires, it
		installs that route, and if it doesn't already have the IGMP Membership
		Request (x,G)
		state for that BD on that ES, it instantiates that IGMP Membership Requ
		est (x,G)
		state. If the DF has not already advertised (originated) a SMET route
		for that (x,G) group in that BD, it does so now.
		</t>
		<t>When the Maximum Response Time timer expires, a PE that has adverti
		sed an
		IGMP Leave Synch route withdraws it. Any PE attached to the
		multihomed ES, which started the Maximum Response Time and has no
		local IGMP Membership Request (x,G) state and no installed IGMP Members
		hip Report
		Synch routes,
		removes the IGMP Membership Request (x,G) state for that (ES,BD). If t
		he DF no
		longer has the IGMP Membership Request (x,G) state for that BD on any E
		S for which it
		is the DF, it withdraws its SMET route for that (x,G) group in that BD.
		</t>
		</section>
		</section>
		<section numbered="true" toc="default">
		<name>Mass Withdraw of the Multicast Membership Report Synch Route in Ca
		se of Failure</name>
		<t>A PE that has received an IGMP Membership Request would have synced t
		he IGMP
		Membership Report by the procedure defined in <xref target="local-igmp-ml
		d"
		format="default"/>. If a PE with the local Membership Report
		state goes down or the PE to CE link goes down, it would lead to a
		mass withdraw of multicast routes. Remote PEs (PEs where these routes
		were remote IGMP Membership Reports) <bcp14>SHOULD NOT</bcp14> remove the
		state immediately;
		instead, General Query <bcp14>SHOULD</bcp14> be generated to refresh the
		states.
		There are several ways to detect failure at a
		peer, e.g., using IGP next-hop tracking or ES route withdraw.
		</t>
		</section>
		</section>
		<section numbered="true" toc="default">
		<name>Single-Active Multihoming</name>
		<t>Note that to facilitate state synchronization after failover, the PEs
		attached to a multihomed ES operating in Single-Active redundancy mode
		<bcp14>SHOULD</bcp14> also coordinate the IGMP Membership Report (x,G) sta
		te.
		In this case, all IGMP
		Membership Report messages are received by the DF and distributed to the n
		on-DF
		PEs using the procedures described above.
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>Selective Multicast Procedures for IR Tunnels</name>
		<t>If an ingress PE uses ingress replication, then for a given (x,G)
		group in a given BD:
		</t>
		<ol spacing="normal" type="1">
		<li>It sends (x,G) traffic to the set of PEs not supporting IGMP or MLD
		Proxies. This set consists of any PE that has advertised an Inclusive Mul
		ticast
		Ethernet Tag (IMET) route for the BD
		without a Multicast Flags extended community or with a Multicast Flags e
		xtended
		community in which neither the IGMP Proxy support nor the MLD Proxy supp
		ort flags are set.
		</li>
		<li>It sends (x,G) traffic to the set of PEs supporting IGMP or MLD Prox
		ies
		and has listeners for that (x,G) group in that BD. This set consists of a
		ny PE
		that has advertised an IMET route for the BD
		with a Multicast Flags extended community in which the IGMP Proxy support
		and/or
		the MLD Proxy support flags are set and that has advertised a SMET route
		for that (x,G)
		group in that BD.
		</li>
		</ol>
		</section>
		<section anchor="bgp-encoding" numbered="true" toc="default">
		<name>BGP Encoding</name>
		<t>This document defines three new BGP EVPN routes to carry IGMP
		Membership Reports. The route types are known as:
		</t>


	<section title="Default Selective Multicast Route">	<dl newline="false" spacing="normal">
	<t>	<dt>6 -</dt>
	If there is multicast router connected behind	<dd>Selective Multicast Ethernet Tag Route </dd>
	the EVPN domain, the PE	<dt>7 -</dt>
	MAY originate a default SMET (,) to get all	<dd>Multicast Membership Report Synch Route </dd>
	multicast traffic in	<dt>8 -</dt>
	domain.	<dd>Multicast Leave Synch Route </dd>
	</t>	</dl>


	<figure >	<t>The detailed encoding and procedures for these route types are
	<artwork ><![CDATA[	described in subsequent sections.
		</t>
		<section anchor="SMET" numbered="true" toc="default">
		<name>Selective Multicast Ethernet Tag Route</name>
		<t>A SMET route-type-specific EVPN NLRI
		consists of the following:
		</t>
		<artwork name="" type="" align="center" alt=""><![CDATA[
		+---------------------------------------+
		\| RD (8 octets) \|
		+---------------------------------------+
		\| Ethernet Tag ID (4 octets) \|
		+---------------------------------------+
		\| Multicast Source Length (1 octet) \|
		+---------------------------------------+
		\| Multicast Source Address (variable) \|
		+---------------------------------------+
		\| Multicast Group Length (1 octet) \|
		+---------------------------------------+
		\| Multicast Group Address (Variable) \|
		+---------------------------------------+
		\| Originator Router Length (1 octet) \|
		+---------------------------------------+
		\| Originator Router Address (variable) \|
		+---------------------------------------+
		\| Flags (1 octet) \|
		+---------------------------------------+
		]]></artwork>
		<t>For the purpose of BGP route key processing, all the fields are
		considered to be part of the prefix in the NLRI, except for the 1-octet
		Flags field. The Flags fields are defined as follows:
		</t>
		<artwork name="" type="" align="center" alt=""><![CDATA[
		0 1 2 3 4 5 6 7
		+--+--+--+--+--+--+--+--+
		\| reserved \|IE\|v3\|v2\|v1\|
		+--+--+--+--+--+--+--+--+
		]]></artwork>
		<ul spacing="normal">
		<li>The least significant bit (bit 7) indicates support for IGMP versi
		on
		1. Since IGMPv1 is being deprecated, the sender <bcp14>MUST</bcp14> set
		it to 0 for IGMP and the receiver <bcp14>MUST</bcp14> ignore it.
		</li>
		<li>The second least significant bit (bit 6) indicates support for IGM
		P
		version 2.
		</li>
		<li>The third least significant bit (bit 5) indicates support for IGMP
		version 3.
		</li>
		<li>The fourth least significant bit (bit 4) indicates whether the (S,
		G)
		information carried within the route-type is of an Include Group type
		(bit value 0) or an Exclude Group type (bit value 1). The Exclude
		Group type bit <bcp14>MUST</bcp14> be ignored if bit 5 is not set.
		</li>
		<li>This EVPN route type is used to carry tenant IGMP multicast group
		information. The Flags field assists in distributing the IGMP Membershi
		p
		Report of a given host for a given multicast route. The version
		bits help associate the IGMP version of receivers participating within
		the EVPN domain.
		</li>
		<li>The IE bit helps in creating filters for a given
		multicast route.
		</li>
		<li>If the route is used for IPv6 (MLD), then bit 7 indicates support
		for MLD
		version 1. The second least significant bit (bit 6) indicates support
		for MLD version 2. Since there is no MLD version 3, in case of IPv6
		routes, the third least significant bit <bcp14>MUST</bcp14> be 0. In ca
		se of IPv6 routes,
		the fourth least significant bit <bcp14>MUST</bcp14> be ignored if bit
		6 is not
		set.
		</li>
		<li> Reserved bits <bcp14>MUST</bcp14> be set to 0 by the sender, and
		the receiver
		<bcp14>MUST</bcp14> ignore the Reserved bits.
		</li>
		</ul>
		<section numbered="true" toc="default">
		<name>Constructing the Selective Multicast Ethernet Tag Route</name>
		<t>This section describes the procedures used to construct the SMET ro
		ute.
		</t>
		<t>The Route Distinguisher (RD) <bcp14>SHOULD</bcp14> be a Type 1 RD <
		xref
		target="RFC4364" format="default"/>. The
		value field comprises an IP address of the PE (typically, the
		loopback address), followed by a number unique to the PE.
		</t>
		<t>The Ethernet Tag ID <bcp14>MUST</bcp14> be set, as per the procedur
		e
		defined in <xref target="RFC7432" format="default"/>.
		</t>
		<t>The Multicast Source Length <bcp14>MUST</bcp14> be set to the lengt
		h of the Multicast
		Source Address in bits. If the Multicast Source Address field
		contains an IPv4 address, then the value of the Multicast Source
		Length field is 32. If the Multicast Source Address field contains an
		IPv6 address, then the value of the Multicast Source Length field is
		128. In case of a (*,G) Membership Report, the Multicast Source Length
		is set to
		0.
		</t>
		<t>The Multicast Source Address is the source IP address from the IGMP
		Membership Report.
		In case of a (*,G) Membership Report, this field is not used.
		</t>
		<t>The Multicast Group Length <bcp14>MUST</bcp14> be set to the length
		of the Multicast Group
		Address in bits. If the Multicast Group Address field contains an
		IPv4 address, then the value of the Multicast Group Length field is
		32. If the Multicast Group Address field contains an IPv6 address,
		then the value of the Multicast Group Length field is 128.
		</t>
		<t>The Multicast Group Address is the group address from the IGMP or M
		LD
		Membership Report.
		</t>
		<t>The Originator Router Length is the length of the Originator Router
		Address in bits.
		</t>
		<t>The Originator Router Address is the IP address of the router origi
		nating this route.
		The SMET Originator Router IP address <bcp14>MUST</bcp14> match that of
		the IMET (or
		S-PMSI Authentic Data (AD))
		route originated for the same EVI by the same downstream PE.
		</t>
		<t>The Flags field indicates the version of IGMP protocol from which t
		he
		Membership Report was received. It also indicates whether the
		multicast group had the INCLUDE or EXCLUDE bit set.
		</t>
		<t> Reserved bits <bcp14>MUST</bcp14> be set to 0. They can be defined
		by other documents in the future. </t>


	+--------------+	<!--[rfced] Does "ToRs" stand for "Top-of-the-Rack (ToR) switches"? If
	\| \|	not, please let us know how to expand this term.
	\| \|
	\| \| +----+
	\| \| \| \|---- H1(*,G1)v2
	\| IP/MPLS \| \| PE1\|---- H2(S2,G2)v3
	\| Network \| \| \|---- S2
	\| \| \| \|
	\| \| +----+
	\| \|
	+----+ \| \|
	+----+ \| \| \| \|
	\| \| S1 ---\| PE2\| \| \|
	\|PIM \|----R1 ---\| \| \| \|
	\|ASM \| +----+ \| \|
	\| \| \| \|
	+----+ +--------------+


	Figure 2: Multicast Router behind EVPN domain	Original:
		IGMP is used to receive group membership information from hosts by
		ToRs.


	]]></artwork>	Perhaps:
	</figure>	IGMP is used to receive group membership information from hosts by
		Top-of-the-Rack (ToR) switches.
		-->


	<t>	<t>IGMP is used to receive group membership information from hosts
	Consider the EVPN network of Figure-2, where there is an EVPN	by ToRs. Upon receiving the host's expression of interest in a
		particular group membership, this information is then forwarded using t
		he
		SMET route. The NLRI also keeps track
		of the receiver's IGMP protocol version and any source filtering for a
		given group membership. All EVPN SMET routes are announced per EVI
		Route Target extended communities (EVI-RT ECs).
		</t>
		</section>
		<section numbered="true" toc="default">
		<name>Reconstructing IGMP/MLD Membership Reports from the Selective Mu
		lticast Route</name>
		<t> This section describes the procedures used to reconstruct IGMP/ML
		D Membership
		Reports from the SMET route.
		</t>
		<ul spacing="normal">
		<li> If the Multicast Group Length is 32, the route would be
		translated to the IGMP membership request. If the Multicast Group
		Length is 128, the route would be translated to an MLD
		membership request. </li>
		<li>The Multicast Group Address field would be translated to
		the IGMP/MLD group address.</li>
		<li> If the Multicast Source Length is set to 0, it would be transla
		ted to any source
		(*).
		If the Multicast Source Length is non-zero, the Multicast Source Add
		ress field would be
		translated to the IGMP/MLD source address.</li>
		<li> If flag bit 7 is set, it translates the Membership report to be
		IGMPv1 or
		MLDv1.</li>
		<li> If flag bit 6 is set, it translates the Membership report to be
		IGMPv2 or MLDv2.</li>
		<li> Flag bit 5 is only valid for the IGMP Membership report; if it i
		s set, it
		translates to the IGMPv3 report.</li>
		<li> If the IE flag is set, it translates to the IGMP/MLD Exclude
		mode membership report. If the IE flag is not set (0), it
		translates to the Include mode membership report. </li>
		</ul>
		</section>
		<section numbered="true" toc="default">
		<name>Default Selective Multicast Route</name>
		<t>If there is a multicast router connected behind the EVPN domain, th
		e PE
		<bcp14>MAY</bcp14> originate a default SMET (,) to get all multicast
		traffic in
		the domain.</t>
		<figure anchor="EVPN-domain">
		<name>Multicast Router behind the EVPN Domain</name>
		<artwork name="" type="" align="center" alt=""><![CDATA[
		+--------------+
		\| \|
		\| \|
		\| \| +----+
		\| \| \| \|---- H1(*,G1)v2
		\| IP/MPLS \| \| PE1\|---- H2(S2,G2)v3
		\| Network \| \| \|---- S2
		\| \| \| \|
		\| \| +----+
		\| \|
		+----+ \| \|
		+----+ \| \| \| \|
		\| \| S1 ---\| PE2\| \| \|
		\|PIM \|----R1 ---\| \| \| \|
		\|ASM \| +----+ \| \|
		\| \| \| \|
		+----+ +--------------+
		]]></artwork>
		</figure>
		<t>Consider the EVPN network in <xref target="EVPN-domain"/>, where th
		ere is an EVPN
	instance configured across the PEs. Let's consider that PE2 is connect ed to	instance configured across the PEs. Let's consider that PE2 is connect ed to
	multicast router R1 and there is a network running PIM ASM behind R1.	multicast router R1 and there is a network running PIM ASM behind R1.

	If there are receivers behind the PIM ASM network the PIM Join would	If there are receivers behind the PIM ASM network, the PIM Join would
	be forwarded to the PIM RP (Rendezvous Point). If receivers behind	be forwarded to the PIM Rendezvous Point (RP). If receivers behind the
	PIM ASM network are interested in a multicast flow originated by	PIM ASM network are interested in a multicast flow originated by
	multicast source S2 (behind PE1), it is necessary for PE2 to receive	multicast source S2 (behind PE1), it is necessary for PE2 to receive

	multicast traffic. In this case PE2 MUST originate a (,) SMET route	multicast traffic. In this case, PE2 <bcp14>MUST</bcp14> originate a ( ,) SMET route
	to receive all of the multicast traffic in the EVPN domain. To generat e	to receive all of the multicast traffic in the EVPN domain. To generat e

	Wildcards (,) routes, the procedure from <xref target="RFC6625"/> MU	wildcard (,) routes, the procedure from <xref target="RFC6625" forma
	ST be used.	t="default"/>
	</t>	<bcp14>MUST</bcp14> be used.</t>
		</section>
	</section>	</section>
		<section numbered="true" toc="default">
	</section>	<name>Multicast Membership Report Synch Route</name>
		<t>This EVPN route type is used to coordinate the IGMP Membership Report
	<section title="Multicast Membership Report Synch Route">	(x,G)
		state for a given BD between the PEs attached to a given ES operating in
	<t>	All-Active (or Single-Active) redundancy mode, and it consists of the
	This EVPN route type is used to coordinate IG	following:</t>
	MP Membership Report (x,G) state for	<artwork name="" type="" align="center" alt=""><![CDATA[+-----------------------
	a given BD between the PEs attached to a give	---------------------------+
	n ES operating in All-	\| RD (8 octets) \|
	Active (or Single-Active) redundancy mode and	+--------------------------------------------------+
	it consists of	\| Ethernet Segment Identifier (10 octets) \|
	following:	+--------------------------------------------------+
	</t>	\| Ethernet Tag ID (4 octets) \|
		+--------------------------------------------------+
	<figure >	\| Multicast Source Length (1 octet) \|
	<artwork ><![CDATA[	+--------------------------------------------------+
		\| Multicast Source Address (variable) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| RD (8 octets) \|	\| Multicast Group Length (1 octet) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| Ethernet Segment Identifier (10 octets) \|	\| Multicast Group Address (Variable) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| Ethernet Tag ID (4 octets) \|	\| Originator Router Length (1 octet) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| Multicast Source Length (1 octet) \|	\| Originator Router Address (variable) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| Multicast Source Address (variable) \|	\| Flags (1 octet) \|
	+--------------------------------------------------+	+--------------------------------------------------+
	\| Multicast Group Length (1 octet) \|	]]></artwork>
	+--------------------------------------------------+	<t>For the purpose of BGP route key processing, all the fields are
	\| Multicast Group Address (Variable) \|	considered to be part of the prefix in the NLRI, except for the 1-octet
	+--------------------------------------------------+	Flags field, whose fields are defined as follows:</t>
	\| Originator Router Length (1 octet) \|	<artwork name="" type="" align="center" alt=""><![CDATA[
	+--------------------------------------------------+	0 1 2 3 4 5 6 7
	\| Originator Router Address (variable) \|	+--+--+--+--+--+--+--+--+
	+--------------------------------------------------+	\| reserved \|IE\|v3\|v2\|v1\|
	\| Flags (1 octet) \|	+--+--+--+--+--+--+--+--+
	+--------------------------------------------------+	]]></artwork>
		<ul spacing="normal">
	]]></artwork>	<li> The least significant bit (bit 7) indicates support for IGMP vers
	</figure>	ion 1. </li>
		<li> The second least significant bit (bit 6) indicates support for IG
	<t>	MP version 2. </li>
	For the purpose of BGP route key processing, a	<li> The third least significant bit (bit 5) indicates support for IGM
	ll the fields are	P version 3. </li>
	considered to be part of the prefix in the NLR	<li> The fourth least significant bit (bit 4) indicates whether the (S
	I except for the one-	, G) information
	octet Flags field, whose fields are defined as	carried within the route-type is of an Include Group type (bit value 0)
	follows:	or an Exclude Group
	</t>	type (bit value 1). The Exclude Group type bit <bcp14>MUST</bcp14> be i
	<figure >	gnored if bit 5 is
	<artwork ><![CDATA[	not set. </li>
		<li> Reserved bits <bcp14>MUST</bcp14> be set to 0.</li>
	0 1 2 3 4 5 6 7	</ul>
	+--+--+--+--+--+--+--+--+	<t>The Flags field assists in distributing the IGMP Membership Report of
	\| reserved \|IE\|v3\|v2\|v1\|	a
	+--+--+--+--+--+--+--+--+	given host for a given multicast route. The version bits help
		associate the IGMP version of receivers participating within the EVPN
	]]></artwork>	domain. The include/exclude bit helps in creating filters for a
	</figure>	given multicast route.</t>
		<t>If the route is being prepared for IPv6 (MLD), then bit 7 indicates
	<t>	support for MLD version 1. The second least significant bit (bit 6)
	<list style="symbols">	indicates support for MLD version 2. Since there is no MLD version 3,
	<t> The least significant bit, bit 7 indicates support	in case of the IPv6 route, the third least significant bit <bcp14>MUST</b
	for IGMP version 1. </t>	cp14>
	<t> The second least significant bit, bit 6 indica	be 0. In case of the IPv6 route, the fourth least significant bit <bcp14>
	tes support for IGMP version 2. </t>	MUST</bcp14>
	<t> The third least significant bit, bit 5 indicat	be ignored if bit 6 is not set.</t>
	es support for IGMP version 3. </t>	<section numbered="true" toc="default">
	<t> The fourth least significant bit, bit 4 indica	<name>Constructing the Multicast Membership Report Synch Route</name>
	tes whether the (S, G) information	<t>This section describes the procedures used to construct the IGMP Me
	carried within the route-type is of Includ	mbership Report
	e Group type (bit value 0) or an Exclude Group type	Synch route. Support for these route types is optional. If a PE does
	(bit value 1). The Exclude Group type bit	not support this route, then it <bcp14>MUST NOT</bcp14> indicate that i
	MUST be ignored if bit 5 is	t supports
	not set. </t>	"IGMP proxy" in the Multicast Flags extended community for the EVIs
	<t> Reserved bits MUST be set to 0.	corresponding to its multihomed ESs.</t>
	</t>	<t>An IGMP Membership Report Synch route <bcp14>MUST</bcp14> carry exa
	</list>	ctly one
		ES-Import Route
	</t>	Target extended community, i.e., the one that corresponds to the ES on
		which the IGMP Membership Report was received. It <bcp14>MUST</bcp14>
	<t>	also carry
	The Flags field assists in distributing IGMP Membership Re	exactly one EVI-RT EC, i.e., the one that corresponds to the EVI on
	port of a	which the IGMP Membership Report
	given host for a given multicast route. The version bits h	was received. See <xref target="evi-rt" format="default"/> for details
	elp	on how to
	associate IGMP version of receivers participating within t	encode and construct the EVI-RT EC.</t>
	he EVPN	<t>The RD <bcp14>SHOULD</bcp14> be Type 1 <xref
	domain. The include/exclude bit helps in creating filters	target="RFC4364" format="default"/>. The
	for a	value field comprises an IP address of the PE (typically, the
	given multicast route.	loopback address), followed by a number unique to the PE.</t>
	</t>	<t>The Ethernet Segment Identifier (ESI) <bcp14>MUST</bcp14> be set to
		the 10-octet
	<t>	value defined for the ES.</t>
	If route is being prepared for IPv6 (MLD) then bit 7	<t>The Ethernet Tag ID <bcp14>MUST</bcp14> be set, as per the procedur
	indicates	e defined in
	support for MLD version 1. The second least signific	<xref target="RFC7432" format="default"/>.</t>
	ant bit, bit 6	<t>The Multicast Source Length <bcp14>MUST</bcp14> be set to the
	indicates support for MLD version 2. Since there is	length of the Multicast Source
	no MLD version 3,	Address in bits. If the Multicast Source field contains an IPv4
	in case of IPv6 route third least significant bit MU	address, then the value of the Multicast Source Length field is 32.
	ST be 0. In case	If the Multicast Source field contains an IPv6 address, then the
	of IPv6 route, the fourth least significant bit MUST	value of the Multicast Source Length field is 128. In case of a (*,G)
	be ignored if	Membership Report, the Multicast Source Length is set to 0.</t>
	bit 6 is not set.	<t>The Multicast Source is the Source IP address of the IGMP Membershi
	</t>	p
		Report. In case of a (*,G) Membership Report, this field does not exis
	<section title="Constructing the Multicast Membership Report S	t.</t>
	ynch Route">	<t>The Multicast Group Length <bcp14>MUST</bcp14> be set to the length
	<t>	of the
	This section describes the procedures used	Multicast Group
	to construct the IGMP Membership Report	Address in bits. If the Multicast Group field contains an IPv4
	Synch route. Support for these route types	address, then the value of the Multicast Group Length field is 32.
	is optional. If a PE does	If the Multicast Group field contains an IPv6 address, then the value
	not support this route, then it MUST NOT in	of the Multicast Group Length field is 128.</t>
	dicate that it supports	<t>The Multicast Group is the group address of the IGMP Membership
	'IGMP proxy' in the Multicast Flag extended	Report.</t>
	community for the EVIs	<t>The Originator Router Length is the length of the Originator Router
	corresponding to its multi-homed Ethernet S	Address in bits.</t>
	egments (ESs).	<t>The Originator Router Address is the IP address of the router origi
	</t>	nating the prefix.</t>
		<t>The Flags field indicates the version of IGMP protocol from which t
	<t>	he
	An IGMP Membership Report Synch route MUST	Membership Report was received. It also indicates whether the
	carry exactly one ES-Import Route	multicast group had the INCLUDE or EXCLUDE bit set.</t>
	Target extended community, the one that cor	<t> Reserved bits <bcp14>MUST</bcp14> be set to 0.</t>
	responds to the ES on	</section>
	which the IGMP Membership Report was receiv	<section numbered="true" toc="default">
	ed. It MUST also carry exactly one	<name>Reconstructing IGMP/MLD Membership Reports from a Multicast Memb
	EVI-RT EC, the one that corresponds to the	ership
	EVI on which the IGMP Membership Report	Report Synch Route</name>
	was received. See <xref target="evi-rt"/>	<t> This section describes the procedures used to reconstruct IGMP/ML
	for details on how to encode and	D
	construct the EVI-RT EC.	Membership Reports from the Multicast Membership Report Synch route.</t
	</t>	>
		<ul spacing="normal">
	<t>	<li> If the Multicast Group Length is 32, the route would be transla
	The Route Distinguisher (RD) SHOULD be a Ty	ted
	pe 1 RD <xref target="RFC4364"/>. The	to the IGMP membership request. If the Multicast Group Length is 128,
	value field comprises an IP address of the	the route would be translated to an MLD membership request. </li>
	PE (typically, the	<li> The Multicast Group Address field would be translated to the
	loopback address) followed by a number uniq	IGMP/MLD group address.</li>
	ue to the PE.	<li> If the Multicast Source Length is set to 0, it would be transla
	</t>	ted to
		any source (*). If the Multicast Source Length is non-zero, the Multi
	<t>	cast Source
	The Ethernet Segment Identifier (ESI) MUST	Address field would be translated to the IGMP/MLD source address.</li
	be set to the 10-octet	>
	value defined for the ES.	<li> If flag bit 7 is set, it translates the Membership report to be
	</t>	IGMPv1
		or MLDv1.</li>
	<t>	<li> If flag bit 6 is set, it translates the Membership report to be
	The Ethernet Tag ID MUST be set as per procedu	IGMPv2
	re defined in <xref target="RFC7432"/>.	or MLDv2.</li>
	</t>	<li> Flag bit 5 is only valid for the IGMP Membership report; if it
		is set,
	<t>	it translates to the IGMPv3 report.</li>
	The Multicast Source length MUST be set to length o	<li> If the IE flag is set, it translates to the IGMP/MLD Exclude mo
	f Multicast Source	de membership
	address in bits. If the Multicast Source field cont	report. If the IE flag is not set (0), it translates to the Include m
	ains an IPv4	ode
	address, then the value of the Multicast Source Len	membership report. </li>
	gth field is 32.	</ul>
	If the Multicast Source field contains an IPv6 addr	</section>
	ess, then the	</section>
	value of the Multicast Source Length field is 128.	<section numbered="true" toc="default">
	In case of a (*,G)	<name>Multicast Leave Synch Route</name>
	Membership Report, the Multicast Source Length is s	<t>This EVPN route type is used to coordinate the IGMP Leave Group (x,G)
	et to 0.	state for a given BD between the PEs attached to a given ES operating
		in an All-Active (or Single-Active) redundancy mode, and it consists of t
	</t>	he
		following:</t>
	<t>	<artwork name="" type="" align="center" alt=""><![CDATA[
	The Multicast Source is the Source IP address	+--------------------------------------------------+
	of the IGMP Membership	\| RD (8 octets) \|
	Report. In case of a (*,G) Membership Report,	+--------------------------------------------------+
	this field does not exist.	\| Ethernet Segment Identifier (10 octets) \|
	</t>	+--------------------------------------------------+
		\| Ethernet Tag ID (4 octets) \|
	<t>	+--------------------------------------------------+
	The Multicast Group length MUST be set to l	\| Multicast Source Length (1 octet) \|
	ength of multicast group	+--------------------------------------------------+
	address in bits. If the Multicast Group fie	\| Multicast Source Address (variable) \|
	ld contains an IPv4	+--------------------------------------------------+
	address, then the value of the Multicast Gr	\| Multicast Group Length (1 octet) \|
	oup Length field is 32.	+--------------------------------------------------+
	If the Multicast Group field contains an IP	\| Multicast Group Address (Variable) \|
	v6 address, then the value	+--------------------------------------------------+
	of the Multicast Group Length field is 128.	\| Originator Router Length (1 octet) \|
	</t>	+--------------------------------------------------+
		\| Originator Router Address (variable) \|
	<t>	+--------------------------------------------------+
	The Multicast Group is the Group address	\| Reserved (4 octets) \|
	of the IGMP Membership	+--------------------------------------------------+
	Report.	\| Maximum Response Time (1 octet) \|
	</t>	+--------------------------------------------------+
		\| Flags (1 octet) \|
	<t>	+--------------------------------------------------+
	The Originator Router Length is the le	]]></artwork>
	ngth of the Originator Router	<t>For the purpose of BGP route key processing, all the fields are
	address in bits.	considered to be part of the prefix in the NLRI, except for the Reserved,
	</t>	Maximum Response Time, and 1-octet Flags fields, which are defined as fol
		lows:</t>
	<t>	<artwork name="" type="" align="center" alt=""><![CDATA[
	The Originator Router Address is the IP add	0 1 2 3 4 5 6 7
	ress of Router Originating the prefix.	+--+--+--+--+--+--+--+--+
	</t>	\| reserved \|IE\|v3\|v2\|v1\|
		+--+--+--+--+--+--+--+--+
	<t>	]]></artwork>
	The Flags field indicates the version of IG	<ul spacing="normal">
	MP protocol from which the	<li> The least significant bit (bit 7) indicates support for IGMP ver
	Membership Report was received. It also ind	sion 1. </li>
	icates whether the	<li> The second least significant bit (bit 6) indicates support for I
	multicast group had INCLUDE or EXCLUDE bit	GMP version 2. </li>
	set.	<li> The third least significant bit (bit 5) indicates support for IG
	</t>	MP version 3. </li>
	<t> Reserved bits MUST be set to	<li> The fourth least significant bit (bit 4) indicates whether the (
	0.	S, G) information
	</t>	carried within the route-type is of an Include Group type (bit value 0)
		or an Exclude Group
	</section>	type (bit value 1). The Exclude Group type bit <bcp14>MUST</bcp14> be
	<section title="Reconstructing I	ignored if bit 5
	GMP / MLD Membership Reports from Multicast Membership Report Sync Route">	is not set. </li>
	<t> Th	<li> Reserved bits <bcp14>MUST</bcp14> be set to 0. They can be define
	is section describes the procedures used to reconstruct IGMP / MLD Membership Re	d by
	ports from Multicast Membership Report Sync route.	other documents in the future. </li>
	</t>	</ul>
	<t>	<t>The Flags field assists in distributing the IGMP Membership Report of
	<list style="symbols">	a
	<t> If multicast group length is 32, rou	given host for a given multicast route. The version bits help
	te would be translated to IGMP membership request. If multicast group length is	associate the IGMP version of the receivers participating within the EVPN
	128, route would be translated to MLD membership request.	domain. The include/exclude bit helps in creating filters for a
	</t>	given multicast route.</t>
		<t>If the route is being prepared for IPv6 (MLD), then bit 7 indicates
	<t> Multicast group address field would be translated	support for MLD version 1. The second least significant bit (bit 6)
	to IGMP / MLD group address.	indicates support for MLD version 2. Since there is no MLD version 3,
	</t>	in case of the IPv6 route, the third least significant bit <bcp14>MUST</b
	<t> If Multicast source length is set to zero it would	cp14> be 0. In case
	be translated to any source (*).	of the IPv6 route, the fourth least significant bit <bcp14>MUST</bcp14> b
	If multicast source length is non zero, Multica	e ignored if
	st source address field would be translated to IGMP / MLD source address.	bit 6 is not set.</t>
	</t>	<t> Reserved bits in the flag <bcp14>MUST</bcp14> be set to 0. They can
	<t> If flag bit 7 is set, it translates Membership repo	be
	rt to be IGMP V1 or MLD V1.	defined by other documents in the future. </t>
	</t>	<section numbered="true" toc="default">
	<t> If flag bit 6 is set, it translates Membership repo	<name>Constructing the Multicast Leave Synch Route</name>
	rt to be IGMP V2 or MLD V2.	<t>This section describes the procedures used to construct the IGMP
	</t>	Leave Synch route. Support for these route types is optional. If a PE
	<t> Flag bit 5 is only valid for IGMP Membership report	does not support this route, then it <bcp14>MUST NOT</bcp14> indicate t
	and if it is set, it translates to IGMP V3 report.	hat it
	</t>	supports "IGMP proxy" in the Multicast Flags extended community for the
	<t> If IE flag is set, it translate to IGMP / MLD Exc	EVIs corresponding to its multihomed Ethernet Segments.</t>
	lude mode membership report. If IE flag is not set (zero), it translates to Incl	<t>An IGMP Leave Synch route <bcp14>MUST</bcp14> carry exactly one ES-
	ude mode membership report.	Import Route
	</t>	Target extended community, i.e., the one that corresponds to the ES on
	</list>	which the IGMP Leave was received. It <bcp14>MUST</bcp14> also carry e
	</t>	xactly one
		EVI-RT EC, i.e., the one that corresponds to the EVI on which the IGMP
	</section>	Leave was received. See <xref target="evi-rt"/> for details on how to
	</section>	form the
		EVI-RT EC.</t>
	<section title="Multicast Leave Synch Route">	<t>The RD <bcp14>SHOULD</bcp14> be Type 1 <xref
	<t>	target="RFC4364" format="default"/>. The
	This EVPN route type is used to coordinate IGMP	value field comprises an IP address of the PE (typically, the
	Leave Group (x,G)	loopback address), followed by a number unique to the PE.</t>
	state for a given BD between the PEs attached to	<t>The ESI <bcp14>MUST</bcp14> be set to the 10-octet
	a given ES operating	value defined for the ES.</t>
	in All-Active (or Single-Active) redundancy mode	<t>The Ethernet Tag ID <bcp14>MUST</bcp14> be set, as per the procedur
	and it consists of	e
	following:	defined in <xref target="RFC7432" format="default"/>.</t>
	</t>	<t>The Multicast Source Length <bcp14>MUST</bcp14> be set to the lengt
		h
	<figure >	of the Multicast Source
	<artwork ><![CDATA[	Address in bits. If the Multicast Source field contains an IPv4
		address, then the value of the Multicast Source Length field is 32.
	+--------------------------------------------------+	If the Multicast Source field contains an IPv6 address, then the
	\| RD (8 octets) \|	value of the Multicast Source Length field is 128. In case of a (*,G)
	+--------------------------------------------------+	Membership Report, the Multicast Source Length is set to 0.</t>
	\| Ethernet Segment Identifier (10 octets) \|	<t>The Multicast Source is the Source IP address of the IGMP Membershi
	+--------------------------------------------------+	p
	\| Ethernet Tag ID (4 octets) \|	Report. In case of a (*,G) Membership Report, this field does not exis
	+--------------------------------------------------+	t.</t>
	\| Multicast Source Length (1 octet) \|	<t>The Multicast Group Length <bcp14>MUST</bcp14> be set to the length
	+--------------------------------------------------+	of the
	\| Multicast Source Address (variable) \|	Multicast Group
	+--------------------------------------------------+	Address in bits. If the Multicast Group field contains an IPv4
	\| Multicast Group Length (1 octet) \|	address, then the value of the Multicast Group Length field is 32.
	+--------------------------------------------------+	If the Multicast Group field contains an IPv6 address, then the value
	\| Multicast Group Address (Variable) \|	of the Multicast Group Length field is 128.</t>
	+--------------------------------------------------+	<t>The Multicast Group is the group address of the IGMP Membership Rep
	\| Originator Router Length (1 octet) \|	ort.</t>
	+--------------------------------------------------+	<t>The Originator Router Length is the length of the Originator Router
	\| Originator Router Address (variable) \|	Address
	+--------------------------------------------------+	in bits.</t>
	\| Reserved (4 octet) \|	<t>The Originator Router Address is the IP address of the router
	+--------------------------------------------------+	originating the prefix.</t>
	\| Maximum Response Time (1 octet) \|	<t> The Reserved field is not part of the route key. The originator <b
	+--------------------------------------------------+	cp14>MUST</bcp14> set
	\| Flags (1 octet) \|	the Reserved field to 0;
	+--------------------------------------------------+	the receiver <bcp14>SHOULD</bcp14> ignore it, and if it needs to be pro
		pagated, it
	]]></artwork>	<bcp14>MUST</bcp14> propagate it unchanged.</t>
	</figure>	<t> The Maximum Response Time is the value to be used while sending a
		query, as defined in
	<t>	<xref target="RFC2236" format="default"/>.</t>
	For the purpose of BGP route key processing, all the fields	<t>The Flags field indicates the version of IGMP protocol from which t
	are	he
	considered to be part of the prefix in the NLRI except for t	Membership Report was received. It also indicates whether the
	he Reserved,	multicast group had an INCLUDE or EXCLUDE bit set.</t>
	Maximum Response Time and the one-octet Flags field, whose f	</section>
	ields are	<section numbered="true" toc="default">
	defined as follows:	<name>Reconstructing IGMP/MLD Leave from a Multicast Leave Synch Route
	</t>	</name>
		<t>This section describes the procedures used to reconstruct IGMP/MLD
	<figure	Leave from
	>	the Multicast Leave Synch route.</t>
	<artwork ><![CDATA[
	0 1 2 3 4 5 6 7
	+--+--+--+--+--+--+--+--+
	\| reserved \|IE\|v3\|v2\|v1\|
	+--+--+--+--+--+--+--+--+

	]]></artwork>
	</figure>

	<t>
	<list style="symbols">
	<t> The least significant bit, bit 7 indicates su
	pport for IGMP version 1. </t>
	<t> The second least significant bit, bit 6 indic
	ates support for IGMP version 2. </t>
	<t> The third least significant bit, bit 5 indica
	tes support for IGMP version 3. </t>
	<t> The fourth least significant bit, bit 4 indic
	ates whether the (S, G) information carried
	within the route-type is of Include Group
	type (bit value 0) or an Exclude Group type
	(bit value 1). The Exclude Group type bit
	MUST be ignored if bit 5 is not set. </t>
	<t> Reserved bits MUST be set to 0. They can b
	e defined in future by other document.
	</t>
	</list>
	</t>

	<t>
	The Flags field assists in distributing IGMP Membershi
	p Report of a
	given host for a given multicast route. The version bi
	ts help
	associate IGMP version of receivers participating with
	in the EVPN
	domain. The include/exclude bit helps in creating fil
	ters for a
	given multicast route.
	</t>

	<t>
	If route is being prepared for IPv6 (MLD) then bit
	7 indicates
	support for MLD version 1. The second least signif
	icant bit, bit 6
	indicates support for MLD version 2. Since there i
	s no MLD version 3,
	in case of IPv6 route third least significant bit
	MUST be 0. In case
	of IPv6 route, the fourth least significant bit M
	UST be ignored if
	bit 6 is not set.
	</t>
	<t> Reserved bits in flag MUST be set to 0. They can be d
	efined in future by other document.
	</t>

	<section title="Constructing the Multicast Leave Synch Ro
	ute">
	<t>
	This section describes the procedures use
	d to construct the IGMP
	Leave Synch route. Support for these ro
	ute types is optional. If a PE
	does not support this route, then it MUS
	T NOT indicate that it
	supports 'IGMP proxy' in Multicast Flag
	extended community for the
	EVIs corresponding to its multi-homed Et
	hernet Segments.
	</t>

	<t>
	An IGMP Leave Synch route MUST carry exact
	ly one ES-Import Route
	Target extended community, the one that co
	rresponds to the ES on
	which the IGMP Leave was received. It MUS
	T also carry exactly one
	EVI-RT EC, the one that corresponds to the
	EVI on which the IGMP
	Leave was received. See Section 9.5 for d
	etails on how to form the
	EVI-RT EC.
	</t>

	<t>
	The Route Distinguisher (RD) SHOULD be
	a Type 1 RD <xref target="RFC4364"/>. The
	value field comprises an IP address of
	the PE (typically, the
	loopback address) followed by a number
	unique to the PE.
	</t>

	<t>
	The Ethernet Segment Identifier (ESI) MUST
	be set to the 10-octet
	value defined for the ES.
	</t>
	<t>
	The Ethernet Tag ID MUST be set as per pr
	ocedure defined in <xref target="RFC7432"/>.
	</t>

	<t>
	The Multicast Source length MUST be set t
	o length of multicast source
	address in bits. If the Multicast Source
	field contains an IPv4
	address, then the value of the Multicast
	Source Length field is 32.
	If the Multicast Source field contains an
	IPv6 address, then the
	value of the Multicast Source Length fiel
	d is 128. In case of a (*,G)
	Membership Report, the Multicast Source
	Length is set to 0.
	</t>

	<t>
	The Multicast Source is the Source IP
	address of the IGMP Membership
	Report. In case of a (*,G) Membership
	Report, this field does not exist.
	</t>

	<t>
	The Multicast Group length MUST be set
	to length of multicast group
	address in bits. If the Multicast Grou
	p field contains an IPv4
	address, then the value of the Multica
	st Group Length field is 32.
	If the Multicast Group field contains
	an IPv6 address, then the value
	of the Multicast Group Length field is
	128.
	</t>

	<t>
	The Multicast Group is the Group addre
	ss of the IGMP Membership Report.
	</t>

	<t>
	The Originator Router Length is the length
	of the Originator Router
	address in bits.
	</t>

	<t>
	The Originator Router Address is the I
	P address of Router Originating the prefix.
	</t>

	<t> Reserved field is not part of the route key.
	The originator MUST set the reserved field to Zero ,
	the receiver SHOULD ignore it and if it n
	eeds to be propagated, it MUST propagate it unchanged
	</t>

	<t> Maximum Response Time is value to be used whi
	le sending query as defined in <xref target="RFC2236"/>
	</t>

	<t>
	The Flags field indicates the version
	of IGMP protocol from which the
	Membership Report was received. It als
	o indicates whether the
	multicast group had INCLUDE or EXCLUDE
	bit set.
	</t>

	</section>

	<section title="Reconstructing IGMP / MLD Leave from Multicast Leave Sync Route"
	>
	<t> Th
	is section describes the procedures used to reconstruct IGMP / MLD Leave from M
	ulticast Leave Sync route.
	</t>
	<t>
	<list style="symbols">
	<t> If multicast group length is 32, rou
	te would be translated to IGMP Leave. If multicast group length is 128, route wo
	uld be translated to MLD Leave.
	</t>

	<t> Multicast group address field would be translated
	to IGMP / MLD group address.
	</t>
	<t> If Multicast source length is set to zero it would
	be translated to any source (*).
	If multicast source length is non zero, Multica
	st source address field would be translated to IGMP / MLD source address.
	</t>
	<t> If flag bit 7 is set, it translates Membership repo
	rt to be IGMP V1 or MLD V1.
	</t>
	<t> If flag bit 6 is set, it translates Membership repo
	rt to be IGMP V2 or MLD V2.
	</t>
	<t> Flag bit 5 is only valid for IGMP Membership report
	and if it is set, it translates to IGMP V3 report.
	</t>
	<t> If IE flag is set, it translate to IGMP / MLD Exc
	lude mode Leave. If IE flag is not set (zero), it translates to Include mode Lea
	ve.
	</t>
	<t>
	</t>
	</list>
	</t>

	</section>

	</section>

	<section title="Multicast Flags Extended Community">
	<t>
	The 'Multicast Flags' extended community is a
	new EVPN extended
	community. EVPN extended communities are tra
	nsitive extended
	communities with a Type field value of 6. IA
	NA will assign a Sub-Type
	from the 'EVPN Extended Community Sub-Types'
	registry.
	</t>

	<t>
	A PE that supports IGMP and/or MLD Proxy on a gi
	ven BD
	MUST attach this extended community to the IMET
	route it advertises
	advertises for that BD and it MUST set the IGMP
	and/or MLD Proxy
	Support flags to 1. Note that an <xref target="R
	FC7432"/> compliant PE will not advertise this
	extended community so its absence indicates that
	the advertising PE
	does not support either IGMP or MLD Proxy.
	</t>

	<t>
	The advertisement of this extended community enab
	les more efficient
	multicast tunnel setup from the source PE special
	ly for ingress
	replication - i.e., if an egress PE supports IGMP
	proxy but doesn't
	have any interest in a given (x,G), it advertises
	its IGMP proxy
	capability using this extended community but it d
	oes not advertise
	any SMET route for that (x,G). When the source PE
	(ingress PE)
	receives such advertisements from the egress PE,
	it does not
	replicate the multicast traffic to that egress PE
	; however, it does
	replicate the multicast traffic to the egress PEs
	that don't
	advertise such capability even if they don't have
	any interests in
	that (x,G).
	</t>

	<t>
	A Multicast Flags extended community is encode
	d as an 8-octet value,
	as follows:
	</t>

	<figure >
	<artwork ><![CDATA[
	0 1 2 3
	0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
	\| Type=0x06 \|Sub-Type=0x09 \| Flags (2 Octets) \|M\|I\|
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
	\| Reserved=0 \|
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

	]]></artwork>
	</figure>

	<t>
	The low-order (lease significant) two bits are defined as th
	e "IGMP
	Proxy Support and MLD Proxy Support" bit. The absence of thi
	s
	extended community also means that the PE does not support I
	GMP
	proxy. where:
	</t>
	<t>
	<list style="symbols">
	<t> Type is 0x06 as registered with IANA for EVPN Exten
	ded Communities. </t>
	<t> Sub-Type : 0x09</t>
	<t>
	Flags are two Octets value.
	<list style="symbols">
	<t> Bit 15 (shown as I) def
	ines IGMP Proxy Support. Value of 1 for
	bit 15 means that P
	E supports IGMP Proxy. Value of 0 for bit 15
	means that PE does
	not supports IGMP Proxy.</t>
	<t>Bit 14 (shown as M) defi
	nes MLD Proxy Support. Value of 1 for
	bit 14 means that
	PE supports MLD Proxy. Value of 0 for bit 14
	means that PE does
	not support MLD proxy. </t>
	<t>Bit 0 to 13 are reserved
	for future. Sender MUST set it 0 and receiver MUST ignore it. </t>
	</list>

	</t>

	<t> Reserved bits are set to 0. Sender MUST set i
	t to 0 and receiver MUST ignore it.</t>
	</list>
	</t>

	<t> If a router does not support this specification, it MUST NOT ad
	d Multicast Flags Extended Community
	in BGP route. A router receiving BGP update,
	if M and I both flag are zero (0), the router MUST treat th
	is Update as malformed. Receiver of such
	update MUST ignore the extended community.
	</t>
	</section>

	<section title="EVI-RT Extended Community" anchor="evi-rt">

	<t>
	In EVPN, every EVI is associated with one
	or more Route Targets
	(RTs). These Route Targets serve two fun
	ctions:
	<list style="numbers">
	<t>
	Distribution control: RTs
	control the distribution of the
	routes. If a route carri
	es the RT associated with a particular
	EVI, it will be distribu
	ted to all the PEs on which that EVI
	exists.
	</t>
	<t>
	EVI identification: O
	nce a route has been received by a
	particular PE, the RT
	is used to identify the EVI to which it
	applies.
	</t>
	</list>

	</t>

	<t>
	An IGMP Membership Report Synch or IGMP L
	eave Synch route is associated with a
	particular combination of ES and EVI. T
	hese routes need to be
	distributed only to PEs that are attache
	d to the associated ES.
	Therefore these routes carry the ES-Imp
	ort RT for that ES.
	</t>

	<t>
	Since an IGMP Membership Report Synch or I
	GMP Leave Synch route does not need
	to be distributed to all the PEs on which
	the associated EVI
	exists, these routes cannot carry the RT a
	ssociated with that
	EVI. Therefore, when such a route arrives
	at a particular PE, the
	route's RTs cannot be used to identify the
	EVI to which the route
	applies. Some other means of associating
	the route with an EVI
	must be used.
	</t>

	<t>
	This document specifies four new Extended Comm
	unities (EC) that
	can be used to identify the EVI with which a
	route is associated,
	but which do not have any effect on the distr
	ibution of the
	route. These new ECs are known as the "Type
	0 EVI-RT EC", the
	"Type 1 EVI-RT EC", the "Type 2 EVI-RT EC", a
	nd the "Type 3 EVI-RT EC".
	<list style="numbers">
	<t> A Type 0 EVI-RT EC is an EVPN EC
	(type 6) of sub-type 0xA.</t>
	<t> A Type 1 EVI-RT EC is an EVPN EC
	(type 6) of sub-type 0xB.</t>
	<t> A Type 2 EVI-RT EC is an EVPN EC
	(type 6) of sub-type 0xC.</t>
	<t> A Type 3 EVI-RT EC is an EVPN EC
	(type 6) of sub-type 0xD</t>
	</list>
	</t>

	<t>
	Each IGMP Membership Report Synch or IGMP Leave S
	ynch route MUST carry exactly
	one EVI-RT EC. The EVI-RT EC carried by a partic
	ular route is
	constructed as follows. Each such route is the r
	esult of having
	received an IGMP Membership Report or an IGMP Lea
	ve message from a particular
	BD. The route is said to be associated with that
	BD.
	For each BD, there is a corresponding RT that is
	used to ensure
	that routes "about" that BD are distributed to al
	l PEs attached
	to that BD. So suppose a given IGMP Membership R
	eport Synch or Leave Synch
	route is associated with a given BD, say BD1, and
	suppose that
	the corresponding RT for BD1 is RT1. Then:

	<list style="symbols">
	<t> 0. If RT1 is a Transitive Two-Octet A
	S-specific EC, then the EVI-
	RT EC carried by the route is a T
	ype 0 EVI-RT EC. The value
	field of the Type 0 EVI-RT EC is
	identical to the value field of
	RT1. </t>


	<t> 1. If RT1 is a Transitive IPv4-Addres	<!--[rfced] In Section 9.3.2, we notice mixed tense in the list of
	s-specific EC, then the EVI-	procedures, i.e., "would be translated" vs. "it translates
	RT EC carried by the route is a T	to". May we update this list to reflect the present tense? Only
	ype 1 EVI-RT EC. The value	the first three bullets would be updated as follows:
	field of the Type 1 EVI-RT EC is
	identical to the value field of
	RT1. </t>
	<t> 2. If RT1 is a Transitive Four-Octet-
	specific EC, then the EVI-RT
	EC carried by the route is a Type
	2 EVI-RT EC. The value field
	of the Type 2 EVI-RT EC is identi
	cal to the value field of RT1.</t>


	<t> 3. If RT1 is a Transitive IPv6-Addres	Original:
	s-specific EC, then the EVI-RT	* If the Multicast Group Length is 32, the route would be translated
	EC carried by the route is a Type	to IGMP Leave. If the Multicast Group Length is 128, the route
	3 EVI-RT EC. The value	would be translated to MLD Leave.
	field of the Type 3 EVI-RT EC is
	identical to the value field of
	RT1.
	</t>


	</list>	* The Multicast Group Address field would be translated to an IGMP/
	</t>	MLD group address.


	<t>	* If the Multicast Source Length is set to 0, it would be translated
	An IGMP Membership Report Synch or Leave S	to any source (*). If the Multicast Source Length is non zero,
	ynch route MUST carry exactly one EVI-RT EC.	the Multicast Source Address field would be translated to the
	</t>	IGMP/MLD source address.


	<t>	Perhaps:
	Suppose a PE receives a particular IGMP Me	* If the Multicast Group Length is 32, the route is translated
	mbership Report Synch or IGMP Leave	to IGMP Leave. If the Multicast Group Length is 128, the route
	Synch route, say R1, and suppose that R1	is translated to MLD Leave.
	carries an ES-Import RT
	that is one of the PE's Import RTs. If
	R1 has no EVI-RT EC, or
	has more than one EVI-RT EC, the PE MUST
	apply the "treat-as-withdraw"
	procedure of <xref target="RFC7606"/>.
	</t>


	<t>	* The Multicast Group Address field is translated to an IGMP/
	Note that an EVI-RT EC is not a Route Targ	MLD group address.
	et Extended Community,
	is not visible to the RT Constrain mechani
	sm <xref target="RFC4684"/>, and is
	not intended to influence the propagation
	of routes by BGP.
	</t>


	<figure >	* If the Multicast Source Length is set to 0, it is translated
	<artwork ><![CDATA[	to any source (*). If the Multicast Source Length is non-zero,
	1 2 3	the Multicast Source Address field is translated to the
	0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1	IGMP/MLD source address.
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+	-->
	\| Type=0x06 \| Sub-Type=n \| RT associated with EVI \|
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
	\| RT associated with the EVI (cont.) \|
	+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+


	]]></artwork>	<ul spacing="normal">
	</figure>	<li> If the Multicast Group Length is 32, the route would be transla
		ted
		to IGMP Leave. If the
		Multicast Group Length is 128, the route would be translated to MLD L
		eave. </li>
		<li> The Multicast Group Address field would be translated to an IGM
		P/MLD group
		address.</li>
		<li> If the Multicast Source Length is set to 0, it would be transla
		ted to
		any source (*).
		If the Multicast Source Length is non-zero, the Multicast Source Add
		ress field would be
		translated to the IGMP/MLD source address.</li>
		<li> If flag bit 7 is set, it translates the Membership report to be
		IGMPv1 or MLDv1.</li>
		<li> If flag bit 6 is set, it translates the Membership report to be
		IGMPv2 or MLDv2.</li>
		<li> Flag bit 5 is only valid for the IGMP Membership report; if it
		is set, it
		translates to the IGMPv3 report.</li>
		<li> If the IE flag is set, it translates to the IGMP/MLD Exclude mo
		de Leave.
		If the IE flag is not set (0), it translates to the Include mode Leav
		e. </li>
		</ul>
		</section>
		</section>
		<section numbered="true" toc="default">
		<name>Multicast Flags Extended Community</name>
		<t>The Multicast Flags extended community is a new EVPN extended
		community. EVPN extended communities are transitive extended
		communities with a Type Value of 0x06. IANA has assigned 0x09 to Multica
		st Flags Extended Community in the "EVPN Extended Community Sub-Types" subregist
		ry.</t>
		<t>A PE that supports IGMP and/or the MLD Proxy on a given BD
		<bcp14>MUST</bcp14> attach this extended community to the IMET route it
		advertises for that BD, and it <bcp14>MUST</bcp14> set the IGMP and/or ML
		D Proxy
		Support flags to 1. Note that a PE compliant with <xref target="RFC7432"
		format="default"/>
		will not advertise this
		extended community, so its absence indicates that the advertising PE
		does not support either IGMP or MLD Proxies.</t>
		<t>The advertisement of this extended community enables a more efficient
		multicast tunnel setup from the source PE specially for ingress
		replication, i.e., if an egress PE supports the IGMP proxy but doesn't
		have any interest in a given (x,G), it advertises its IGMP proxy
		capability using this extended community, but it does not advertise
		any SMET route for that (x,G). When the source PE (ingress PE)
		receives such advertisements from the egress PE, it does not
		replicate the multicast traffic to that egress PE; however, it does
		replicate the multicast traffic to the egress PEs that don't
		advertise such capability, even if they don't have any interests in
		that (x,G).</t>
		<t>A Multicast Flags extended community is encoded as an 8-octet value
		as follows:</t>
		<artwork name="" type="" align="center" alt=""><![CDATA[
		0 1 2 3
		0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		\| Type=0x06 \|Sub-Type=0x09 \| Flags (2 Octets) \|M\|I\|
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		\| Reserved=0 \|
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		]]></artwork>


	<t>	<t>The low-order (least significant) 2 bits are defined as the "IGMP
	Where the value of 'n' is 0x0A, 0x0B, 0x0C, or 0x0D corr	Proxy Support" and "MLD Proxy Support" bits (see <xref target="multicast_
	esponding	flags_extended_community"/>. The absence of this
	to EVI-RT type 0, 1, 2, or 3 respectively.	extended community also means that the PE does not support the IGMP
	</t>	proxy, where:</t>
		<ul spacing="normal">
		<li> The Type is 0x06, as registered with IANA for EVPN Extended Commu
		nities. </li>
		<li> The Sub-Type is 0x09.</li>
		<li>
		<t>Flags are 2-octet values.</t>


	</section>	<ul spacing="normal">
		<li> Bit 15 (shown as I) defines IGMP Proxy Support. The value of
		1 for
		bit 15 means that the PE supports the IGMP Proxy. The value of 0 fo
		r bit 15
		means that the PE does not support the IGMP Proxy.</li>
		<li>Bit 14 (shown as M) defines MLD Proxy Support. The value of 1
		for
		bit 14 means that the PE supports the MLD Proxy. The value of 0 for
		bit 14
		means that the PE does not support the MLD proxy. </li>
		<li>Bits 0 to 13 are reserved for the future. The sender <bcp14>MU
		ST</bcp14>
		set it to 0, and the receiver <bcp14>MUST</bcp14> ignore it. </li>
		</ul>
		</li>
		<li> Reserved bits are set to 0. The sender <bcp14>MUST</bcp14> set it
		to 0,
		and the receiver <bcp14>MUST</bcp14> ignore it.</li>
		</ul>
		<t> If a router does not support this specification, it <bcp14>MUST NOT<
		/bcp14> add
		the Multicast Flags Extended Community
		in the BGP route. When a router receives a BGP update,
		if both M and I flags are 0, the router <bcp14>MUST</bcp14> treat this up
		date as
		malformed. The receiver of such an
		update <bcp14>MUST</bcp14> ignore the extended community. </t>
		</section>
		<section anchor="evi-rt" numbered="true" toc="default">
		<name>EVI-RT Extended Community</name>
		<t>In EVPN, every EVI is associated with one or more Route Targets. The
		se RTs serve two functions:</t>
		<ol spacing="normal" type="1">
		<li>Distribution control: RTs control the distribution of the
		routes. If a route carries the RT associated with a particular
		EVI, it will be distributed to all the PEs on which that EVI
		exists.</li>
		<li>EVI identification: Once a route has been received by a
		particular PE, the RT is used to identify the EVI to which it
		applies.</li>
		</ol>
		<t>An IGMP Membership Report Synch or IGMP Leave Synch route is associat
		ed with a
		particular combination of ES and EVI. These routes need to be
		distributed only to PEs that are attached to the associated ES.
		Therefore, these routes carry the ES-Import RT for that ES.</t>
		<t>Since an IGMP Membership Report Synch or IGMP Leave Synch route does
		not need
		to be distributed to all the PEs on which the associated EVI
		exists, these routes cannot carry the RT associated with that
		EVI. Therefore, when such a route arrives at a particular PE, the
		route's RTs cannot be used to identify the EVI to which the route
		applies. Some other means of associating the route with an EVI
		must be used.</t>
		<t>This document specifies four new ECs that
		can be used to identify the EVI with which a route is associated
		but do not have any effect on the distribution of the
		route. These new ECs are known as "Type 0 EVI-RT EC",
		"Type 1 EVI-RT EC", "Type 2 EVI-RT EC", and "Type 3 EVI-RT EC".</t>


	<section title="Rewriting of RT ECs and EVI-RT ECs by ASBRs">	<ol spacing="normal" type="1">
	<t>	<li> A Type 0 EVI-RT EC is an EVPN EC (type 6) of sub-type 0xA.</li>
	There are certain situations in which an	<li> A Type 1 EVI-RT EC is an EVPN EC (type 6) of sub-type 0xB.</li>
	ES is attached to a set	<li> A Type 2 EVI-RT EC is an EVPN EC (type 6) of sub-type 0xC.</li>
	of PEs that are not all in the same AS, o	<li> A Type 3 EVI-RT EC is an EVPN EC (type 6) of sub-type 0xD</li>
	r not all operated by	</ol>
	the same provider. In some such situati	<t>Each IGMP Membership Report Synch or IGMP Leave Synch route <bcp14>MU
	ons, the RT that	ST</bcp14>
	corresponds to a particular EVI may be d	carry exactly
	ifferent in each AS. If	one EVI-RT EC. The EVI-RT EC carried by a particular route is
	a route is propagated from AS1 to AS2, a	constructed as follows. Each such route is the result of having
	n ASBR at the AS1/AS2	received an IGMP Membership Report or an IGMP Leave message from a partic
	border may be provisioned with a policy	ular
	that removes the RTs that	BD. The route is said to be associated with that BD.
	are meaningful in AS1 and replaces them	For each BD, there is a corresponding RT that is used to ensure
	with the corresponding	that routes "about" that BD are distributed to all PEs attached
	(i.e., RTs corresponding to the same EVI	to that BD. So suppose a given IGMP Membership Report Synch or Leave Syn
	s) RTs that are	ch
	meaningful in AS2. This is known as RT-	route is associated with a given BD, say BD1, and suppose that
	rewriting.	the corresponding RT for BD1 is RT1. Then:</t>
	</t>


	<t>	<!--[rfced] For consistency, we updated "Two-Octet" and "Four-Octet"
	Note that if a given route's RTs are rewri	to "2-octet" and "4-octet". Please let us know of any
	tten, and the route	objections.
	carries an EVI-RT EC, the EVI-RT EC needs
	to be rewritten as
	well.
	</t>
	</section>


	<section title="BGP Error Handling">	Original:
	<t>	* If RT1 is a Transitive Two-Octet AS-specific EC, then the EVI-RT
	If a received BGP update contains Flags not in a	EC carried by the route is a Type 0 EVI-RT EC.
	ccordance with IGMP/MLD version-X expectation,
	the PE MUST apply the "treat-as-withdraw" proced
	ure as per <xref target="RFC7606"/>
	</t>
	<t>
	If a received BGP update is malformed such that
	BGP route keys cannot be extracted, then
	BGP update MUST be considered as invalid. Receiv
	ing PE MUST apply the "Session reset" procedure of <xref target="RFC7606"/>.
	</t>


	</section>	* If RT1 is a Transitive Four-Octet-specific EC, then the EVI-RT EC
		carried by the route is a Type 2 EVI-RT EC.


	</section>	Perhaps:
		* If RT1 is a Transitive 2-octet AS-specific EC, then the EVI-RT
		EC carried by the route is a Type 0 EVI-RT EC.


	<section title="IGMP Version 1 Membership Report">	* If RT1 is a Transitive 4-octet-specific EC, then the EVI-RT EC
	<t>	carried by the route is a Type 2 EVI-RT EC.
	This document does not provide any detail about IGMPv1 p	-->
	rocessing.	<ul spacing="normal">
	Implementations are expected to only use IGMPv2 and above for IPv4 and	<li>If RT1 is a Transitive 2-octet AS-specific EC, then the EVI-RT
	MLDv1 and above for IPv6. IGMPv1 routes are considered invalid and the	EC carried by the route is a Type 0 EVI-RT EC. The value
	PE MUST apply the "treat-as-withdraw" procedure as per <xref target="RFC7606"	field of the Type 0 EVI-RT EC is identical to the value field of
	/>.	RT1. </li>
		<li>If RT1 is a Transitive IPv4-Address-specific EC, then the EVI-RT
		EC carried by the route is a Type 1 EVI-RT EC. The value
		field of the Type 1 EVI-RT EC is identical to the value field of
		RT1. </li>
		<li>If RT1 is a Transitive 4-octet-specific EC, then the EVI-RT
		EC carried by the route is a Type 2 EVI-RT EC. The value field
		of the Type 2 EVI-RT EC is identical to the value field of RT1.</li>
		<li>If RT1 is a Transitive IPv6-Address-specific EC, then the EVI-RT
		EC carried by the route is a Type 3 EVI-RT EC. The value
		field of the Type 3 EVI-RT EC is identical to the value field of
		RT1.</li>
		</ul>
		<t>An IGMP Membership Report Synch or Leave Synch route <bcp14>MUST</bcp
		14>
		carry exactly one EVI-RT EC.</t>
		<t>Suppose a PE receives a particular IGMP Membership Report Synch or IG
		MP Leave
		Synch route, say R1, and suppose that R1 carries an ES-Import RT
		that is one of the PE's Import RTs. If R1 has no EVI-RT EC or
		has more than one EVI-RT EC, the PE <bcp14>MUST</bcp14> apply the "treat-
		as-withdraw"
		procedure per <xref target="RFC7606" format="default"/>.</t>
		<t>Note that an EVI-RT EC is not a Route Target extended community,
		is not visible to the RT Constrain mechanism <xref target="RFC4684" forma
		t="default"/>,
		and is not intended to influence the propagation of routes by BGP.</t>
		<artwork name="" type="" align="center" alt=""><![CDATA[
		1 2 3
		0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		\| Type=0x06 \| Sub-Type=n \| RT associated with EVI \|
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		\| RT associated with the EVI (cont.) \|
		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
		]]></artwork>


	</t>	<t>The value of "n" is 0x0A, 0x0B, 0x0C, or 0x0D, corresponding
	</section>	to EVI-RT types 0, 1, 2, or 3, respectively.</t>
		</section>
		<section numbered="true" toc="default">
		<name>Rewriting of RT ECs and EVI-RT ECs by ASBRs</name>
		<!-- Note: text updated per mail from John E Drake <jdrake@juniper.net>
		on 5/18/2022. -->


	<section title="Security Considerations">	<!-- [rfced] AD, Section 9.6 was updated per a request from the author. Please
	<t>	review and let us know if the changes are approved. Note that you can also view
	This document describes a means to efficiently operate IGMP and ML	the changes in the diff files.
	D on a subnet
	constructed across multiple PODs or DCs via an EVPN solution. The
	security
	considerations for the operation of the underlying EVPN and BGP su
	bstrate are
	described in <xref target="RFC7432"/>, and specific multicast cons
	iderations are outlined in
	<xref target="RFC6513"/> and <xref target="RFC6514"/>. The EVPN a
	nd associated IGMP proxy provides a single
	broadcast domain so the same security considerations of IGMPv2 <xr
	ef target="RFC2236"/>,
	<xref target="RFC3376"/>, MLD <xref target="RFC2710"/>, or MLDv2 <
	xref target="RFC3810"/> apply.


	</t>	Original:
	</section>	There are certain situations in which an ES is attached to a set of
		PEs that are not all in the same AS, or not all operated by the same
		provider. In some such situations, the RT that corresponds to a
		particular EVI may be different in each AS. If a route is propagated
		from AS1 to AS2, an ASBR at the AS1/AS2 border may be provisioned
		with a policy that removes the RTs that are meaningful in AS1 and
		replaces them with the corresponding (i.e., RTs corresponding to the
		same EVIs) RTs that are meaningful in AS2. This is known as RT-
		rewriting.


	<section title="IANA Considerations">	Note that if a given route's RTs are rewritten, and the route carries
		an EVI-RT EC, the EVI-RT EC needs to be rewritten as well.


	<section title="EVPN Extended Community Sub-Types Registrations">	Current:
	<t>	There are certain situations in which an ES is attached to a set of
	IANA has allocated the following codepoints from the EVPN Extended Co	PEs that are not all in the same AS, or not all operated by the same
	mmunity Sub-Types	provider. In this situation, the RT that corresponds to a particular
	sub-registry of the BGP Extended Communities registry. </t>	EVI may be different in each AS. If a route is propagated from AS1
		to AS2, an ASBR at the AS1/AS2 border may be configured with a policy
		that replaces the EVI RTs for AS1 with the corresponding EVI RTs
		for AS2. This is known as RT-rewriting.


	<figure>	If an ASBR is configured to perform RT-rewriting of the EVI RTs in
	<artwork ><![CDATA[	EVPN routes, it MUST be configured to perform RT-rewriting of the
		corresponding EVI-RT extended communities in IGMP Join Synch and IGMP
		Leave Synch Routes.
		-->


	0x09 Multicast Flags Extended Community [this document]	<t>There are certain situations in which an ES is attached to a set of P
	0x0A EVI-RT Type 0 [this document]	Es that are not all in the same AS, or not all operated by the same provider. I
	0x0B EVI-RT Type 1 [this document]	n this situation, the RT that corresponds to a particular EVI may be different i
	0x0C EVI-RT Type 2 [this document]	n each AS. If a route is propagated from AS1 to AS2, an ASBR at the AS1/AS2 bor
		der may be configured with a policy that replaces the EVI RTs for AS1 with the c
		orresponding EVI RTs for AS2. This is known
		as RT-rewriting.</t>
		<t>If an ASBR is configured to perform RT-rewriting of the EVI RTs in EV
		PN routes, it <bcp14>MUST</bcp14> be configured to perform RT-rewriting of the c
		orresponding EVI-RT extended communities in IGMP Join Synch and IGMP Leave Sync
		h Routes.</t>
		</section>
		<section numbered="true" toc="default">
		<name>BGP Error Handling</name>
		<t>If a received BGP update contains Flags not in accordance with the IG
		MP/MLD
		version-X expectation,
		the PE <bcp14>MUST</bcp14> apply the "treat-as-withdraw" procedure per <x
		ref
		target="RFC7606" format="default"/>.</t>
		<t>If a received BGP update is malformed such that BGP route keys cannot
		be extracted,
		then the BGP update <bcp14>MUST</bcp14> be considered invalid. The receiv
		ing PE
		<bcp14>MUST</bcp14> apply the "session reset" procedure per <xref target=
		"RFC7606"
		format="default"/>.</t>
		</section>
		</section>
		<section numbered="true" toc="default">
		<name>IGMP Version 1 Membership Report</name>
		<t>This document does not provide any detail about IGMPv1 processing.
		Implementations are expected to only use IGMPv2 and above for IPv4 and
		MLDv1 and above for IPv6. IGMPv1 routes are considered invalid, and the
		PE <bcp14>MUST</bcp14> apply the "treat-as-withdraw" procedure per
		<xref target="RFC7606" format="default"/>.</t>
		</section>
		<section numbered="true" toc="default">
		<name>Security Considerations</name>
		<t>This document describes a means to efficiently operate IGMP and MLD on
		a subnet
		constructed across multiple PODs or DCs via an EVPN solution. The securit
		y
		considerations for the operation of the underlying EVPN and BGP substrates
		are
		described in <xref target="RFC7432" format="default"/>, and specific multi
		cast
		considerations are outlined in
		<xref target="RFC6513" format="default"/> and <xref target="RFC6514" forma
		t="default"/>.


	]]></artwork>	<!--[rfced] As RFC 3376 specifies IGMPv3, may we update this sentence as follows
	</figure>	?


	<t>	Original:
	IANA is requested to allocate a new codepoint from the EVP	The EVPN and associated IGMP proxy provides a single
	N	broadcast domain so the same security considerations of IGMPv2
	Extended Community sub-types registry for the following.	[RFC2236], [RFC3376], MLD [RFC2710], or MLDv2 [RFC3810] apply.
	</t>


	<figure>	Perhaps:
	<artwork ><![CDATA[	The EVPN and associated IGMP proxy provides a single
	0x0D EVI-RT Type 3 [this document]	broadcast domain so the same security considerations of IGMPv2
		[RFC2236], IGMPv3 [RFC3376], MLD [RFC2710], or MLDv2 [RFC3810] apply.
		-->


	]]></artwork>	The EVPN and associated IGMP proxy provides a single
	</figure>	broadcast domain so the same security considerations of IGMPv2 <xref targe
	</section>	t="RFC2236"
	<section title="EVPN Route Type Registration">	format="default"/>
		<xref target="RFC3376" format="default"/>, MLD <xref target="RFC2710" form
		at="default"/>,
		or MLDv2 <xref target="RFC3810" format="default"/> apply.</t>
		</section>


	<t>	<section numbered="true" toc="default">
	IANA has allocated the following EVPN route types from the EVPN	<name>IANA Considerations</name>
	Route Type registry.	<section numbered="true" toc="default">
	</t>	<name>EVPN Extended Community Sub-Types Registration</name>
		<t>IANA has allocated the following codepoints in the "EVPN Extended Com
		munity Sub-Types"
		subregistry under the "Border Gateway Protocol (BGP) Extended Communitie
		s" registry. </t>


	<figure>	<table>
	<artwork ><![CDATA[	<name>EVPN Extended Community Sub-Types Subregistry Allocated Codepoint
	6 - Selective Multicast Ethernet Tag Route	s</name>
	7 - Multicast Membership Report Synch Route	<thead>
	8 - Multicast Leave Synch Route	<tr>
		<th>Sub-Type Value</th>
		<th>Name</th>
		<th>Reference</th>
		</tr>
		</thead>
		<tbody>
		<tr>
		<td>0x09</td>
		<td>Multicast Flags Extended Community</td>
		<td>RFC 9251</td>
		</tr>
		<tr>
		<td>0x0A</td>
		<td>EVI-RT Type 0</td>
		<td>RFC 9251</td>
		</tr>
		<tr>
		<td>0x0B</td>
		<td>EVI-RT Type 1</td>
		<td>RFC 9251</td>
		</tr>
		<tr>
		<td>0x0C</td>
		<td>EVI-RT Type 2</td>
		<td>RFC 9251</td>
		</tr>
		<tr>
		<td>0x0D</td>
		<td>EVI-RT Type 3</td>
		<td>RFC 9251</td>
		</tr>
		</tbody>
		</table>
		</section>
		<section numbered="true" toc="default">
		<name>EVPN Route Types Registration</name>
		<t>IANA has allocated the following EVPN route types in the "EVPN
		Route Types" subregistry.</t>


	]]></artwork>	<dl newline="false" spacing="normal">
	</figure>	<dt>6 -</dt>
	</section>	<dd>Selective Multicast Ethernet Tag Route</dd>
	<section title="Multicast Flags Extended Community Registry">	<dt>7 -</dt>
		<dd>Multicast Membership Report Synch Route</dd>
		<dt>8 -</dt>
		<dd> Multicast Leave Synch Route</dd>
		</dl>


	<t>	</section>
	The Multicast Flags Extended Community contains a 16-bit Fl	<section anchor="multicast_flags_extended_community" numbered="true" toc="
	ags	default">
	field. The bits are numbered 0-15, from high-order to low-o	<name>Multicast Flags Extended Community Registry</name>
	rder.	<t>IANA has created and now maintains a new subregistry called "Multicas
	</t>	t Flags Extended Community" under the "Border Gateway Protocol (BGP) Extended Co
		mmunities" registry. The registration procedure is First Come First Served <xref
		target="RFC8126"/>. For the 16-bit Flags field, the bits are numbered 0-15, fro
		m high order to low order. The registry was initialized as follows:</t>


	<figure>	<table>
	<artwork ><![CDATA[	<name>Multicast Flags Extended Community</name>
	The registry should be initialized as follows:	<thead>
		<tr>
		<th>Bit</th>
		<th>Name</th>
		<th>Reference</th>
		<th>Change Controller</th>
		</tr>
		</thead>


	Bit Name Reference Change	<tbody>
	Controller	<tr>
	---- -------------- ------------- -------	<td>0-13</td>
	-----------	<td>Unassigned</td>
	0 - 13 Unassigned	<td></td>
	14 MLD Proxy Support This document. IE	<td></td>
	TF	</tr>
	15 IGMP Proxy Support This document IE
	TF


	The registration policy should be "First Come First Served".	<tr>
		<td> 14</td>
		<td>MLD Proxy Support</td>
		<td>RFC 9251</td>
		<td>IETF</td>
		</tr>


	]]></artwork>	<tr>
	</figure>	<td> 15</td>
	</section>	<td>IGMP Proxy Support</td>
		<td>RFC 9251</td>
		<td>IETF</td>
		</tr>
		</tbody>
		</table>
		</section>
		</section>
		</middle>
		<back>
		<displayreference target="I-D.ietf-bess-evpn-bum-procedure-updates" to="EVPN
		-BUM"/>
		<references>
		<name>References</name>
		<references>
		<name>Normative References</name>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.2119.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.7432.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.3376.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.2710.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.3810.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.7606.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.4684.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.2236.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.8174.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.4364.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.6625.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.6513.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.6514.xml"/>
		</references>
		<references>
		<name>Informative References</name>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.R
		FC.4541.xml"/>
		<xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RF
		C.8126.xml"/>


	</section>	<!-- draft-ietf-bess-evpn-bum-procedure-updates-14: in MISSREF as of 5/26
		/22-->
		<xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D
		.ietf-bess-evpn-bum-procedure-updates.xml"/>
		</references>
		</references>
		<section numbered="false" toc="default">
		<name>Acknowledgements</name>
		<t>The authors would like to thank <contact fullname="Stephane Litkowski"/
		>, <contact
		fullname="Jorge Rabadan"/>,
		<contact fullname="Anoop Ghanwani"/>, <contact fullname="Jeffrey Haas"/>,
		<contact
		fullname="Krishna Muddenahally Ananthamurthy"/>, and <contact fullname="Sw
		adesh Agrawal"/>
		for their reviews and valuable comments.</t>
		</section>
		<section numbered="false" toc="default">
		<name>Contributors</name>


	<section title="Acknowledgement">	<contact fullname="Derek Yeung">
	<t>	<organization>Arrcus</organization>
	The authors would like to thank Stephane Litkowski, Jor	<address>
	ge Rabadan,	<email>derek@arrcus.com</email>
	Anoop Ghanwani, Jeffrey Haas, Krishna Muddenahally Anan	</address>
	thamurthy, Swadesh Agrawal	</contact>
	for reviewing and providing valuable
	comment.
	</t>


	</section>	</section>
		</back>


	<section title="Contributors">	<!--[rfced] Please review the "Inclusive Language" portion of the
		online Style Guide
		<https://www.rfc-editor.org/styleguide/part2/#inclusive_language>
		and let us know if any changes are needed. Note that our script
		did not flag any words in particular, but this should still be
		reviewed as a best practice. -->


	<t>	<!--[rfced] Throughout the text, the following terminology appears to be used in
	Derek Yeung </t>	consistently. Please review these occurrences and let us know if/how these shoul
	<t> Arrcus </t>	d be made consistent.
	<t> Email: derek@arrcus.com
	</t>
	</section>


	</middle>	- Ethernet Segment vs. Ethernet segment


	<!-- ***BACK MATTER *** -->	- EVPN Extended Community vs. EVPN extended community
		[Note that RFC 7432 uses "EVPN Extended Community" (capitalized)]


	<back>	- IGMP/MLD Proxy vs. IGMP/MLD proxy
	<references title='Normative References'>	- IGMP Report vs. IGMP report
	<?rfc include='reference.RFC.2119' ?>	- include/exclude bit vs. INCLUDE or EXCLUDE bit
	<?rfc include='reference.RFC.7432' ?>	- Membership Report vs. Membership report vs. membership report
	<?rfc include='reference.RFC.3376' ?>	- Membership Queries vs. membership queries
	<?rfc include='reference.RFC.2710' ?>	- Membership Request vs. membership request
	<?rfc include='reference.RFC.3810' ?>
	<?rfc include='reference.RFC.7606' ?>
	<?rfc include='reference.RFC.4684' ?>
	<?rfc include='reference.RFC.2236' ?>
	<?rfc include='reference.RFC.8174' ?>
	<?rfc include='reference.RFC.4364' ?>
	<?rfc include='reference.RFC.6625' ?>
	<?rfc include='reference.RFC.6513' ?>
	<?rfc include='reference.RFC.6514' ?>


	</references>	- Multicast Flags extended community vs. Multicast Flags Extended Community
	<references title="Informative References">	[Note that we recommend using "Multicast Flags Extended Community" (capitalize
	<?rfc include='reference.RFC.4541' ?>	d)
	<?rfc include="reference.I-D.ietf-bess-evpn-bum-procedure-updates"?>	to match the IANA registry]
	</references>


	</back>	- proxy reporting vs. Proxy-reporting
		- route type vs. route-type
		- Source IP address vs. source IP address
		-->
	</rfc>	</rfc>

End of changes. 103 change blocks.
	2513 lines changed or deleted	2041 lines changed or added
This html diff was produced by rfcdiff 1.48. The latest version is available from http://tools.ietf.org/tools/rfcdiff/

mirror server hosted at Truenetwork, Russian Federation.