MP2MP Ingress Repl.: Simulating Partial Mesh of Multipoint-to-Multipoint (MP2MP) Provider Tunnels with Ingress Replication
Carrying bidirectional customer multicast over a provider network using ingress replication instead of MP2MP tunnels.
Zhang, Z., Rekhter, Y., and A. Dolganow, "Simulating Partial Mesh of Multipoint-to-Multipoint (MP2MP) Provider Tunnels with Ingress Replication", RFC 7740, DOI 10.17487/RFC7740, January 2016, <https://www.rfc-editor.org/info/rfc7740>.
@misc{rfc7740,
series = {Request for Comments},
number = 7740,
howpublished = {RFC 7740},
publisher = {RFC Editor},
doi = {10.17487/RFC7740},
url = {https://www.rfc-editor.org/info/rfc7740},
author = {Z. Zhang and Y. Rekhter and A. Dolganow},
title = {{Simulating Partial Mesh of Multipoint-to-Multipoint (MP2MP) Provider Tunnels with Ingress Replication}},
pagetotal = 8,
year = 2016,
month = jan,
} RFC text
Contents
Internet Engineering Task Force (IETF) Z. Zhang
Request for Comments: 7740 Y. Rekhter
Category: Standards Track Juniper Networks
ISSN: 2070-1721 A. Dolganow
Alcatel-Lucent
January 2016
Simulating Partial Mesh of Multipoint-to-Multipoint (MP2MP)
Provider Tunnels with Ingress Replication
Abstract
RFC 6513 ("Multicast in MPLS/BGP IP VPNs") describes a method to
support bidirectional customer multicast flows using a partial mesh
of Multipoint-to-Multipoint (MP2MP) tunnels. This document specifies
how a partial mesh of MP2MP tunnels can be simulated using Ingress
Replication. This solution enables a service provider to use Ingress
Replication to offer transparent bidirectional multicast service to
its VPN customers.
Status of This Memo
This is an Internet Standards Track document.
This document is a product of the Internet Engineering Task Force
(IETF). It represents the consensus of the IETF community. It has
received public review and has been approved for publication by the
Internet Engineering Steering Group (IESG). Further information on
Internet Standards is available in Section 2 of RFC 5741.
Information about the current status of this document, any errata,
and how to provide feedback on it may be obtained at
http://www.rfc-editor.org/info/rfc7740.
Copyright Notice
Copyright (c) 2016 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . 3
1.2. Requirements Language . . . . . . . . . . . . . . . . . . 4
2. Operation . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1. Control State . . . . . . . . . . . . . . . . . . . . . . 4
2.2. Forwarding State . . . . . . . . . . . . . . . . . . . . 6
3. Security Considerations . . . . . . . . . . . . . . . . . . . 7
4. References . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.1. Normative References . . . . . . . . . . . . . . . . . . 7
4.2. Informative References . . . . . . . . . . . . . . . . . 8
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . 8
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 8
1. Introduction
Section 11.2 of RFC 6513 ("Partitioned Sets of PEs") describes two
methods of carrying Bidirectional PIM (BIDIR-PIM) [RFC5015] C-flow
traffic over a provider core without using the core as the Rendezvous
Point Link (RPL) or requiring Designated Forwarder election.
With these two methods, all Provider Edges (PEs) of a particular VPN
are separated into partitions, with each partition being all the PEs
that elect the same PE as the Upstream PE with respect to the C-RPA
(the Rendezvous Point Address in the customer's address space). A PE
must discard bidirectional C-flow traffic from PEs that are not in
the same partition as the PE itself.
In particular, Section 11.2.3 of RFC 6513 ("Partial Mesh of MP2MP
P-Tunnels") guarantees the above discard behavior without using an
extra PE Distinguisher Label by having all PEs in the same partition
join a single MP2MP tunnel dedicated to that partition and use it to
transmit traffic. All traffic arriving on the tunnel will be from
PEs in the same partition, so it will be always accepted.
RFC 6514 specifies BGP encodings and procedures used to implement
Multicast VPN (MVPN) as specified in RFC 6513, while the details
related to MP2MP tunnels are specified in [RFC7582].
RFC 7582 assumes that an MP2MP P-tunnel is realized either via BIDIR-
PIM [RFC5015] or via MP2MP mLDP (Multipoint extensions for LDP)
[RFC6388]. Each would require signaling and state not just on PEs,
but on the P routers as well. This document describes how the MP2MP
tunnel can be simulated with a mesh of P2MP tunnels, each of which is
instantiated by Ingress Replication (IR) [RFC6513] [RFC6514]. The
procedures in this document are different from the procedures that
are used to set up the mesh of Ingress Replication tunnels as
described in RFC 6514; the procedures in this document do not require
each PE on the MP2MP tunnel to send a Selective P-Multicast Service
Interface (S-PMSI) auto-discovery route (A-D route) for the P2MP
tunnel that the PE is the root for, nor do they require each PE to
send a Leaf A-D route to the root of each P2MP tunnel in the mesh.
Because it uses Ingress Replication, this scheme has both the
advantages and the disadvantages of Ingress Replication in general.
1.1. Terminology
This document uses terminology from [RFC5015], [RFC6513], [RFC6514],
and [RFC7582].
1.2. Requirements Language
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in [RFC2119].
2. Operation
In the following sections, the originator of an S-PMSI A-D route or
Leaf A-D route is determined from the "originating router's IP
address" field of the corresponding route.
2.1. Control State
If a PE, say PEx, is connected to a site of a given VPN and PEx's
next-hop interface to some C-RPA is a VPN Routing and Forwarding
(VRF) interface, then PEx MUST advertise a (C-*,C-*-BIDIR) S-PMSI A-D
route, regardless of whether it has any local BIDIR-PIM join states
corresponding to the C-RPA learned from its Customer Edges (CEs). It
MAY also advertise one or more (C-*,C-G-BIDIR) S-PMSI A-D routes, if
selective distribution trees are needed for those C-G-BIDIR groups
and the corresponding C-RPA is in the site that the PEx connects to.
For example, the (C-*,C-G-BIDIR) S-PMSI A-D routes could be triggered
when the (C-*,C-G-BIDIR) traffic rate goes above a threshold (this
may require measuring the traffic in both directions, due to the
nature of BIDIR-PIM), and fan-out could also be taken into account.
The S-PMSI A-D routes include a PMSI Tunnel Attribute (PTA) with
tunnel type set to Ingress Replication, with the Leaf Information
Required flag set, with a downstream allocated MPLS label that other
PEs in the same partition MUST use when sending relevant C-BIDIR
flows to this PE, and with the Tunnel Identifier field in the PTA set
to a routable address of the originator. This specification does not
prevent sharing of labels between P-tunnels, such as a label being
shared by a (C-*,C-*-BIDIR) and a (C-*,C-G-BIDIR) S-PMSI A-D route
originated by a given PE (note that other specifications put
constraints on how that can be done, e.g., [MVPN-EXTRANET]).
If some other PE, PEy, receives and imports into one of its VRFs any
(C-*,C-*-BIDIR) S-PMSI A-D route whose PTA specifies an IR P-tunnel
and the VRF has any local BIDIR-PIM join state that PEy has received
from its CEs and if PEy chooses PEx as its Upstream PE with respect
to the C-RPA for those states, PEy MUST advertise a Leaf A-D route in
response. Or, if PEy has received and imported into one of its VRFs
a (C-*,C-*-BIDIR) S-PMSI A-D route from PEx before, then upon
receiving in the VRF any local BIDIR-PIM join state from its CEs with
PEx being the Upstream PE for those states' C-RPA, PEy MUST advertise
a Leaf A-D route.
The encoding of the Leaf A-D route is as specified in RFC 6514,
except that the Route Targets are set to the same value as in the
corresponding S-PMSI A-D route so that the Leaf A-D route will be
imported by all VRFs that import the corresponding S-PMSI A-D route.
This is irrespective of whether or not the originator of the S-PMSI
A-D route is the Upstream PE from a receiving PE's perspective. The
label in the PTA of the Leaf A-D route originated by PEy MUST be
allocated specifically for PEx, so that when traffic arrives with
that label, the traffic can associate with the partition (represented
by the PEx). This specification does not prevent sharing of labels
between P-tunnels, such as a label being shared by a (C-*,C-*-BIDIR)
and a (C-*,C-G-BIDIR) Leaf A-D route originated by a given PE (note
that other specifications put constraints on how that can be done,
e.g., [MVPN-EXTRANET]).
Note that RFC 6514 requires that a PE or an ASBR (Autonomous System
Border Router) take no action with regard to a Leaf A-D route unless
that Leaf A-D route carries an IP-address-specific Route Target
identifying the PE/ASBR. This document removes that requirement when
the route key of a Leaf A-D route identifies a (C-*,C-*-BIDIR) or a
(C-*,C-G-BIDIR) S-PMSI.
To speed up convergence (so that PEy starts receiving traffic from
its new Upstream PE immediately instead of waiting until the new Leaf
A-D route corresponding to the new Upstream PE is received by sending
PEs), PEy MAY advertise a Leaf A-D route even if it does not choose
PEx as its Upstream PE with respect to the C-RPA. With that, it will
receive traffic from all PEs, but some will arrive with the label
corresponding to its choice of Upstream PE while some will arrive
with a different label; the traffic in the latter case will be
discarded.
Similar to the (C-*,C-*-BIDIR) case, if PEy receives and imports into
one of its VRFs any (C-*,C-G-BIDIR) S-PMSI A-D route whose PTA
specifies an IR P-tunnel, PEy chooses PEx as its Upstream PE with
respect to the C-RPA, and it has corresponding local (C-*,C-G-BIDIR)
join state that it has received from its CEs in the VRF, PEy MUST
advertise a Leaf A-D route in response. Or, if PEy has received and
imported into one of its VRFs a (C-*,C-G-BIDIR) S-PMSI A-D route
before, then upon receiving its local (C-*,C-G-BIDIR) join state from
its CEs in the VRF, it MUST advertise a Leaf A-D route.
The encoding of the Leaf A-D route is similar to the (C-*,C-*-BIDIR)
case. Similarly, PEy MAY advertise a Leaf A-D route even if it does
not choose PEx as its Upstream PE with respect to the C-RPA.
PEy MUST withdraw the corresponding Leaf A-D route if any of the
following conditions are true:
o the (C-*,C-*-BIDIR) or (C-*,C-G-BIDIR) S-PMSI A-D route is
withdrawn.
o PEy no longer chooses the originator PEx as its Upstream PE with
respect to C-RPA and PEy only advertises Leaf A-D routes in
response to its Upstream PE's S-PMSI A-D route.
o if relevant local join state is pruned.
2.2. Forwarding State
The specification regarding forwarding state in this section matches
the "When an S-PMSI is a 'Match for Transmission'" and "When an
S-PMSI is a 'Match for Reception'" rules for the "Flat Partitioning"
method in [RFC7582], except that the rules about (C-*,C-*) are not
applicable, because this document requires that (C-*,C-*-BIDIR)
S-PMSI A-D routes are always originated for a VPN that supports
C-BIDIR flows.
For the (C-*,C-G-BIDIR) S-PMSI A-D route that a PEy receives and
imports into one of its VRFs from its Upstream PE with respect to the
C-RPA, if PEy itself advertises the S-PMSI A-D route in the VRF, PEy
maintains a (C-*,C-G-BIDIR) forwarding state in the VRF, with the
Ingress Replication provider tunnel leaves being the originators of
the S-PMSI A-D route and all relevant Leaf A-D routes. The relevant
Leaf A-D routes are the routes whose Route Key field contains the
same information as the MCAST-VPN Network Layer Reachability
Information (NLRI) of the (C-*,C-G-BIDIR) S-PMSI A-D route advertised
by the Upstream PE.
For the (C-*,C-*-BIDIR) S-PMSI A-D route that a PEy receives and
imports into one of its VRFs from its Upstream PE with respect to a
C-RPA, if PEy itself advertises the S-PMSI A-D route in the VRF, it
maintains appropriate forwarding states in the VRF for the ranges of
bidirectional groups for which the C-RPA is responsible. The
provider tunnel leaves are the originators of the S-PMSI A-D route
and all relevant Leaf A-D routes. The relevant Leaf A-D routes are
the routes whose Route Key field contains the same information as the
MCAST-VPN NLRI of the (C-*,C-*-BIDIR) S-PMSI A-D route advertised by
the Upstream PE. This is for the so-called "Sender Only Branches"
where a router only has data to send upstream towards C-RPA but no
explicit join state for a particular bidirectional group. Note that
the traffic must be sent to all PEs (not just the Upstream PE) in the
partition, because they may have specific (C-*,C-G-BIDIR) join states
that this PEy is not aware of, while there are no corresponding
(C-*,C-G-BIDIR) S-PMSI A-D and Leaf A-D routes.
For a (C-*,C-G-BIDIR) join state that a PEy has received from its CEs
in a VRF, if there is no corresponding (C-*,C-G-BIDIR) S-PMSI A-D
route from its Upstream PE in the VRF, PEy maintains a corresponding
forwarding state in the VRF, with the provider tunnel leaves being
the originators of the (C-*,C-*-BIDIR) S-PMSI A-D route and all
relevant Leaf A-D routes (same as the "Sender Only Branches" case
above). The relevant Leaf A-D routes are the routes whose Route Key
field contains the same information as the MCAST-VPN NLRI of the
(C-*,C-*-BIDIR) S-PMSI A-D route originated by the Upstream PE. If
there is also no (C-*,C-*-BIDIR) S-PMSI A-D route from its Upstream
PE, then the provider tunnel has an empty set of leaves, and PEy does
not forward relevant traffic across the provider network.
3. Security Considerations
This document raises no new security issues. Security considerations
for the base protocol are covered in [RFC6513] and [RFC6514].
4. References
4.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<http://www.rfc-editor.org/info/rfc2119>.
[RFC6513] Rosen, E., Ed. and R. Aggarwal, Ed., "Multicast in MPLS/
BGP IP VPNs", RFC 6513, DOI 10.17487/RFC6513, February
2012, <http://www.rfc-editor.org/info/rfc6513>.
[RFC6514] Aggarwal, R., Rosen, E., Morin, T., and Y. Rekhter, "BGP
Encodings and Procedures for Multicast in MPLS/BGP IP
VPNs", RFC 6514, DOI 10.17487/RFC6514, February 2012,
<http://www.rfc-editor.org/info/rfc6514>.
[RFC7582] Rosen, E., Wijnands, IJ., Cai, Y., and A. Boers,
"Multicast Virtual Private Network (MVPN): Using
Bidirectional P-Tunnels", RFC 7582, DOI 10.17487/RFC7582,
July 2015, <http://www.rfc-editor.org/info/rfc7582>.
4.2. Informative References
[MVPN-EXTRANET]
Rekhter, Y., Ed., Rosen, E., Ed., Aggarwal, R., Cai, Y.,
and T. Morin, "Extranet Multicast in BGP/IP MPLS VPNs",
Work in Progress, draft-ietf-bess-mvpn-extranet-06,
January 2016.
[RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano,
"Bidirectional Protocol Independent Multicast (BIDIR-
PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007,
<http://www.rfc-editor.org/info/rfc5015>.
[RFC6388] Wijnands, IJ., Ed., Minei, I., Ed., Kompella, K., and B.
Thomas, "Label Distribution Protocol Extensions for Point-
to-Multipoint and Multipoint-to-Multipoint Label Switched
Paths", RFC 6388, DOI 10.17487/RFC6388, November 2011,
<http://www.rfc-editor.org/info/rfc6388>.
Acknowledgements
We would like to thank Eric Rosen for his comments and suggestions
for some text used in the document.
Authors' Addresses
Zhaohui Zhang
Juniper Networks
10 Technology Park Dr.
Westford, MA 01886
United States
Email: zzhang@juniper.net
Yakov Rekhter
Juniper Networks
Andrew Dolganow
Alcatel-Lucent
600 March Rd.
Ottawa, ON K2K 2E6
Canada
Email: andrew.dolganow@alcatel-lucent.com
AI ELI5
On this page
§1The Problem and the Approach
Section 11.2 of RFC 6513 describes two methods of carrying Bidirectional PIM (BIDIR-PIM) customer traffic over a provider core without using the core as the Rendezvous Point Link or requiring Designated Forwarder election. All PEs of a VPN are split into partitions: each partition is the set of PEs that choose the same PE as Upstream PE with respect to the C-RPA, the Rendezvous Point Address in the customer's address space. A PE must discard bidirectional traffic from PEs outside its own partition.
RFC 7582 assumes an MP2MP P-tunnel is built with BIDIR-PIM (RFC 5015) or MP2MP mLDP (RFC 6388). Each needs signaling and state on P routers too. This document instead simulates the MP2MP tunnel with a mesh of P2MP tunnels, each instantiated by Ingress Replication. Because of that, the scheme inherits the advantages and the disadvantages of Ingress Replication in general.
The old way builds a special shared route through the whole network. The new way has each sender just make copies for the other members of its group, so the network in the middle needs no special setup.
§1.1How It Differs from RFC 6514 Ingress Replication
The procedures here differ from those that set up the mesh of Ingress Replication tunnels in RFC 6514. Each PE on the MP2MP tunnel does not need to send a Selective P-Multicast Service Interface (S-PMSI) auto-discovery route (A-D route) for the P2MP tunnel it roots, and it does not need to send a Leaf A-D route to the root of every P2MP tunnel in the mesh.
Terminology comes from RFC 5015, RFC 6513, RFC 6514 and RFC 7582.
This works a little differently from the usual copy-and-send method, and it skips some of the announcements that method needs.
§2.1Advertising S-PMSI A-D Routes
A PE (PEx) connected to a VPN site, whose next-hop interface toward some C-RPA is a VRF interface, MUST advertise a (C-*,C-*-BIDIR) S-PMSI A-D route, even if it has no local BIDIR-PIM join state for that C-RPA. It MAY also advertise (C-*,C-G-BIDIR) S-PMSI A-D routes when selective trees are needed for particular groups. The RFC suggests these could be triggered when the traffic rate goes above a threshold, which may require measuring both directions, and fan-out could also be considered.
These routes carry a PMSI Tunnel Attribute (PTA) with the tunnel type set to Ingress Replication, the Leaf Information Required flag set, a downstream-allocated MPLS label that other PEs in the same partition MUST use when sending relevant flows to this PE, and a Tunnel Identifier set to a routable address of the originator. Labels may be shared between P-tunnels.
Each member of the group announces: "I am here, and here is the special ticket number to put on packages for me."
§2.1Responding with Leaf A-D Routes
When another PE (PEy) imports a (C-*,C-*-BIDIR) S-PMSI A-D route whose PTA specifies an Ingress Replication tunnel, has local BIDIR-PIM join state from its CEs, and chooses PEx as its Upstream PE for the C-RPA, PEy MUST advertise a Leaf A-D route. The same holds if it already imported the route and the join state arrives later. For (C-*,C-G-BIDIR) routes the rule is parallel.
The Leaf A-D route is encoded as in RFC 6514, except that its Route Targets equal those of the matching S-PMSI A-D route, so all VRFs importing that route import the Leaf A-D route too. The label in its PTA MUST be allocated specifically for PEx, so arriving traffic with that label can be tied to the partition PEx represents. This document also removes the RFC 6514 requirement that a PE take no action on a Leaf A-D route lacking an IP-address-specific Route Target, for these two route kinds.
To speed convergence, PEy MAY advertise a Leaf A-D route even if it does not choose PEx as Upstream PE. It then receives traffic from all PEs, and traffic with a label for a different choice is discarded.
PEy MUST withdraw the Leaf A-D route if the S-PMSI A-D route is withdrawn, if PEy no longer chooses PEx as Upstream PE (and only advertises in response to its Upstream PE), or if the relevant local join state is pruned.
When a member wants to join in, it replies with its own ticket number. The ticket tells the sender which group it belongs to, so packages from other groups get thrown away.
- S-PMSI A-D routePTA: Ingress Replication, Leaf Information Required, MPLS label.
- Leaf A-D routeLabel allocated specifically for PEx.
§2.2Forwarding State
The forwarding rules match the "Match for Transmission" and "Match for Reception" rules of the "Flat Partitioning" method in RFC 7582, except that the (C-*,C-*) rules do not apply, since (C-*,C-*-BIDIR) S-PMSI A-D routes are always originated for a VPN supporting C-BIDIR flows.
If PEy itself advertises the S-PMSI A-D route and imports one from its Upstream PE, it keeps forwarding state whose Ingress Replication leaves are the originators of that route and of all relevant Leaf A-D routes, meaning those whose Route Key matches the Upstream PE's NLRI.
For "Sender Only Branches", where a router only sends upstream toward the C-RPA with no explicit join state, traffic must go to all PEs in the partition, not just the Upstream PE, because they may hold (C-*,C-G-BIDIR) joins that PEy does not know about. If there is also no (C-*,C-*-BIDIR) route from the Upstream PE, the tunnel has an empty set of leaves and PEy does not forward the traffic across the provider network.
A sender copies its package to every member it knows about. If it knows nobody, it sends nothing at all.
§3Security Considerations
The document raises no new security issues. Security considerations for the base protocol are covered in RFC 6513 and RFC 6514.
Nothing new to worry about here beyond what the earlier rules already cover.