One document matched: draft-ietf-l3vpn-mvpn-considerations-03.xml
<?xml version="1.0"?>
<?rfc toc="yes"?><?rfc tocompact="yes"?><?rfc tocdepth="3"?><?rfc tocindent="yes"?><?rfc symrefs="yes"?><?rfc sortrefs="no"?><?rfc comments="yes"?><?rfc inline="yes"?><?rfc compact="yes"?><?rfc subcompact="no"?><rfc category="info" docName="draft-ietf-l3vpn-mvpn-considerations-03" ipr="trust200811">
<front>
<title abbrev="Multicast VPN mandatory features">Mandatory Features in a
Layer 3 Multicast BGP/MPLS VPN Solution</title>
<author fullname="Thomas Morin" initials="T." role="editor" surname="Morin">
<organization>France Telecom R&D</organization>
<address>
<postal>
<street>2 rue Pierre Marzin</street>
<city>Lannion</city>
<code>22307</code>
<country>France</country>
</postal>
<email>thomas.morin@orange-ftgroup.com</email>
</address>
</author>
<author fullname="Ben Niven-Jenkins" initials="B.P." role="editor" surname="Niven-Jenkins">
<organization>BT</organization>
<address>
<postal>
<street>208 Callisto House, Adastral Park</street>
<city>Ipswich</city>
<region>Suffolk</region>
<code>IP5 3RE</code>
<country>UK</country>
</postal>
<email>benjamin.niven-jenkins@bt.com</email>
</address>
</author>
<author fullname="Yuji Kamite" initials="Y." surname="Kamite">
<organization abbrev="NTT Communications">NTT Communications
Corporation</organization>
<address>
<postal>
<street>Tokyo Opera City Tower</street>
<street>3-20-2 Nishi Shinjuku, Shinjuku-ku</street>
<region>Tokyo</region>
<code>163-1421</code>
<country>Japan</country>
</postal>
<email>y.kamite@ntt.com</email>
</address>
</author>
<author fullname="Raymond Zhang" initials="R." surname="Zhang">
<organization>BT</organization>
<address>
<postal>
<street>2160 E. Grand Ave.</street>
<city>El Segundo</city>
<code>CA 90025</code>
<country>USA</country>
</postal>
<email>raymond.zhang@bt.com</email>
</address>
</author>
<author fullname="Nicolai Leymann" initials="N." surname="Leymann">
<organization>Deutsche Telekom</organization>
<address>
<postal>
<street>Goslarer Ufer 35</street>
<city>10589 Berlin</city>
<country>Germany</country>
</postal>
<email>n.leymann@telekom.de</email>
</address>
</author>
<author fullname="Nabil Bitar" initials="N" surname="Bitar">
<organization>Verizon</organization>
<address>
<postal>
<street>40 Sylvan Road</street>
<city>Waltham</city>
<region>MA</region>
<code>02451</code>
<country>USA</country>
</postal>
<email>nabil.n.bitar@verizon.com</email>
</address>
</author>
<date day="29" month="April" year="2009"/>
<abstract>
<t>More that one set of mechanisms to support multicast in a layer 3
BGP/MPLS VPN has been defined. These are presented in the documents that
define them as optional building blocks.</t>
<t>To enable interoperability between implementations, this document
defines a subset of features that is considered mandatory for a
multicast BGP/MPLS VPN implementation. This will help implementers and
deployers understand which L3VPN multicast requirements are best
satisfied by each option.</t>
</abstract>
<note title="Requirements Language">
<t>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 <xref target="RFC2119"/>.</t>
</note>
</front>
<middle>
<section title="Introduction">
<t>Specifications for <xref target="I-D.ietf-l3vpn-2547bis-mcast">multicast in BGP/MPLS</xref>
include multiple alternative mechanisms for some of the required
building blocks of the solution. However, they do not identify which of
these mechanisms are mandatory to implement in order to ensure
interoperability. Not defining a set of mandatory to implement
mechanisms leads to a situation where implementations may support
different subsets of the available optional mechanisms which do not
interoperate, which is a problem for the numerous operators having
multi-vendor backbones.</t>
<t>The aim of this document is to leverage the already expressed <xref target="RFC4834">requirements</xref> and study the properties of each
approach, to identify mechanisms that are good candidates for being part
of a core set of mandatory mechanisms which can be used to provide a
base for interoperable solutions.</t>
<t>This document goes through the different building blocks of the
solution and concludes on which mechanisms an implementation is required
to implement. <xref target="summary"/> summarizes these
requirements.</t>
<t>Considering the history of the multicast VPN proposals and
implementations, it is also useful to discuss how existing deployments
of early implementations <xref target="I-D.rosen-vpn-mcast"/><xref target="I-D.raggarwa-l3vpn-2547-mvpn"/> can be accommodated, and
provide suggestions in this respect.</t>
</section>
<section title="Terminology">
<t>Please refer to <xref target="I-D.ietf-l3vpn-2547bis-mcast"/>
and <xref target="RFC4834"/>.</t>
</section>
<section title="Examining alternatives mechanisms for MVPN functions">
<section anchor="autodiscovery" title="MVPN auto-discovery">
<t><xref target="I-D.ietf-l3vpn-2547bis-mcast">The current solution
document</xref> proposes two different mechanisms for MVPN
auto-discovery:<list style="numbers">
<t>BGP-based auto-discovery</t>
<t>"PIM/shared tree" : discovery done through the exchange of PIM
Hellos by C-PIM instances, across an MI-PMSI implemented with one
shared tree per VPN (using multicast ASM, or MP2MP LDP)</t>
</list>Both solutions address <xref target="RFC4834">Section 5.2.10
of</xref> which states that "the operation of a multicast VPN solution
SHALL be as light as possible and providing automatic configuration
and discovery SHOULD be a priority when designing a multicast VPN
solution. Particularly the operational burden of setting up multicast
on a PE or for a VR/VRF SHOULD be as low as possible".</t>
<t>The key consideration is that PIM-based discovery is only
applicable to deployments using a shared tree to instantiate an
MI-PMSI (it cannot be applicable to if only P2P or SSM trees are used,
because contrary to ASM and MP2MP, building these P2P or SSM trees
cannot happen before the autodiscovery has been done), whereas the
BGP-based auto-discovery does not place any constraint on the type of
multicast trees that would have to be used. BGP-based auto-discovery
is independent of the type of P-multicast tree used thus satisfying
the requirement in <xref target="RFC4834">section 5.2.4.1 of</xref>
that "a multicast VPN solution SHOULD be designed so that control and
forwarding planes are not interdependent".</t>
<t>Additionally, it is to be noted that a number of service providers
have chosen to use SSM-based trees for the default MDTs within their
current deployments, therefore relying already on some BGP-based
auto-discovery.</t>
<t>Moreover, when shared P-tunnels are used, the use of BGP
auto-discovery would allow inconsistencies in the
addresses/identifiers used for the shared trees to be detected (e.g.
the same shared tree identifier being used for different VPNs with
distinct BGP route targets). This is particularly attractive in the
context of inter-AS VPNs where the impact of any misconfiguration
could be magnified and where a single service provider may not operate
all the ASs. Note that this technique to detect some misconfiguration
cases may not be usable during a transition period from a shared-tree
autodiscovery to a BGP-based autodiscovery.</t>
<t>Thus, the recommendation is that implementation of the BGP-based
auto-discovery is mandated and should be supported by all mVPN
implementations.</t>
</section>
<section title="S-PMSI Signaling">
<t><xref target="I-D.ietf-l3vpn-2547bis-mcast">The current solution
document</xref> proposes two mechanisms for signaling that multicast
flows will be switched to an S-PMSI :</t>
<t><list style="numbers">
<t>a UDP-based TLV protocol specifically for S-PMSI signaling
(described in section 7.4.2).</t>
<t>a BGP-based mechanism for S-PMSI signaling (described in
section 7.4.1).</t>
</list></t>
<t><xref target="RFC4834">Section 5.2.10 of</xref> states that "as far
as possible, the design of a solution SHOULD carefully consider the
number of protocols within the core network: if any additional
protocols are introduced compared with the unicast VPN service, the
balance between their advantage and operational burden SHOULD be
examined thoroughly". The UDP-based mechanism would be an additional
protocol in the MVPN stack, which isn't the case for the BGP-based
S-PMSI switching signaling, since (a) BGP is identified as a
requirement for autodiscovery, and (b) the BGP-based S-PMSI switching
signaling procedures are very similar to the autodiscovery
procedures.</t>
<t>Furthermore, the BGP-based S-PMSI switching signaling mechanism can
be used within MVPNs using either a UI-PMSI or a MI-PMSI while the
UDP-based protocol is restricted to use within MVPNs using an MI-PMSI.
In practice, this means that, except if shared trees are used, a PE
will have to join to all trees of all PEs in a VPN, while in the
alternative where BGP-based S-PMSI switching signaling is used, it
could delay joining a tree from a PE until traffic from that PE is
needed, thus reducing the amount of state maintained on P routers.</t>
<t>S-PMSI switching signaling approaches can also be compared in an
inter-AS context (see <xref target="interas"/>). The proposed
BGP-based approach for S-PMSI switching signaling provides a good fit
with both the segmented and non-segmented inter-AS approaches
(see<xref target="interas"/>). By contrast the UDP-based
approach for S-PMSI switching signaling appears to be usable with
segmented inter-AS tunnels, but in that case key advantages of the
segmented approach are lost :</t>
<t><list style="symbols">
<t>there is no more an independence of ASes to choose when S-PMSIs
tunnels will be triggered in their AS (and thus control the amount
of state created on their P routers), and with which tunneling
technique they will be built</t>
<t>in an inter-AS option B context, an isolation of ASes is
obtained as PEs don't have visibility of, nor exchange with, PEs
of other ASes. This property can be preserved if the segmented
inter-AS approach and BGP-based S-PMSI switching signaling are
used, but it is not preserved if UDP-based switching signaling is
used.</t>
</list></t>
<t>Given all the above, it is the recommendation of the authors that
BGP is the preferred solution for S-PMSI switching signaling and
should be supported by all implementations.</t>
<t>It is identified that, if nothing prevents a fast-paced creation of
S-PMSI, then S-PMSI switching signaling with BGP would possibly impact
the Route Reflectors used for mVPN routes. However is it also
identified that such a fast-paced behavior would have an impact on P
and PE routers resulting from S-PMSI tunnels signaling, which will be
the same independently of the S-PMSI signaling approach that is used,
and which it is certainly best to avoid by setting up proper
mechanisms.</t>
<t>The UDP-based S-PMSI switching signaling protocol can also be
considered, as an option, given that this protocol has been in
deployment for some time. Implementations supporting both protocols
would be expected to provide a per-VRF configuration knob to allow an
implementation to use the UDP-based TLV protocol for S-PMSI switching
signaling for specific VRFs in order to support the coexistence of
both protocols (for example during migration scenarios). Apart from
such migration-facilitating mechanisms, the authors specifically do
not recommend extending the already proposed UDP-based TLV protocol to
new types of P-multicast trees.</t>
</section>
<section title="PE-PE Transmission of C-Multicast Routing">
<t><xref target="I-D.ietf-l3vpn-2547bis-mcast">The current solution
document</xref> proposes multiple mechanisms for PE-PE transmission of
customer multicast routing information:<list style="numbers">
<t>Full per-MVPN PIM peering across an MI-PMSI (described in
section 3.4.1.1).</t>
<t>Lightweight PIM peering across an MI-PMSI (described in section
3.4.1.2)</t>
<t>The unicasting of PIM C-Join/Prune messages (described in
section 3.4.1.3)</t>
<t>The use of BGP for carrying C-Multicast routing (described in
section 3.4.2).</t>
</list></t>
<section anchor="pepe-scaling" title="PE-PE signaling scalability">
<t>Scalability being one of the core requirements for multicast VPN,
it is useful to compare the proposed C-multicast routing mechanisms
from this perspective : <xref target="RFC4834">Section 4.2.4
of</xref> recommends that "a multicast VPN solution SHOULD support
several hundreds of PEs per multicast VPN, and MAY usefully scale up
to thousands" and section 4.2.5 states that "a solution SHOULD scale
up to thousands of PEs having multicast service enabled".</t>
<t>Scalability with an increased number of VPNs per PE, or with an
increased number of multicast state per VPN, are also important, but
are not focused on in this section since we didn't identify
differences between the different approaches for these matters : all
others things equal, the load on PE due to C-multicast routing
increases roughly linearly with the number of VPNs per PE, and with
the number of multicast state per VPN.</t>
<t>This section presents conclusions related to PE-PE signaling
scalability, based on <xref target="PEPE-mrouting-load"/> that
provides more detailed explanations on the differences in ways of
handling the C-multicast routing load, between the PIM-based
approaches and the BGP-based approach, along with a quantified
evaluations of the amount of state and messages with the different
approaches. Many points made in this section are base on the
conclusions in <xref target="quant-conclusion"/>.</t>
<t>At high scales of multicast deployment, the first and third
mechanisms require the PEs to maintain a large number of PIM
adjacencies with other PEs of the same multicast VPN (which implies
the regular exchange PIM Hellos with each other) and to refresh
C-Join/Prune states, resulting in an increased processing cost when
the amount of PEs increases (as detailed in <xref target="PEPE-mrouting-load"/>) to which the second approach is
less subject, and to which the fourth approach is not subject.</t>
<t>The third mechanism would reduce the amount of C-Join/Prune
processing for a given multicast flow for PEs that are not the
upstream neighbor for this flow, but would require "explicit
tracking" state to be maintained by the upstream PE. It also isn't
compatible with the "Join suppression" mechanism. A possible way to
reduce the amount of signaling with this approach would be the use
of a PIM refresh-reduction mechanism. Such a mechanism, based on
TCP, is being specified by the PIM IETF Working Group (<xref target="I-D.ietf-pim-port"/>) ; its use in a multicast VPN
context has not been described in <xref target="I-D.ietf-l3vpn-2547bis-mcast"/>, but it is expected
that this approach would provide a scalability similar with the
BGP-based approach without RR.</t>
<t>The second mechanism would operate in a similar manner to full
per-MVPN PIM peering except that PIM Hello messages are not
transmitted and PIM C-Join/Prune refresh-reduction would be used,
thereby improving scalability, but this approach has yet to be fully
described. In any case, it seems that it only improves one thing
among the things that will impact scalability with an increased
number of PEs.</t>
<t>The first and second mechanisms can leverage the "Join
suppression" behavior and thus improve the processing burden of an
upstream PE, sparing the processing of a Join refresh message for
each remote PE joined to a multicast stream. This improvement
requires all PEs of a multicast VPN to process all PIM Join and
Prune messages sent by any other PE participating in the same
multicast VPN whether they are the upstream PE or not.</t>
<t>The fourth mechanism (the use of BGP for carrying C-Multicast
routing) would have a comparable drawback of requiring all PEs to
process a BGP C-multicast route only interesting a specific upstream
PE. For this reason the C-multicast routing approach can leverage
the Route-Target constraint mechanisms, which specifically allows
only the interested upstream PE to receive a BGP C-multicast route.
When RT-constraints are used the fourth mechanism reduces the total
amount of message processing load put on the PEs for customer
multicast routing to the minimum (by avoiding any processing by
"unrelated" PEs, that are not the joining PE nor the upstream PE,
and by avoiding the use of refreshes), and inherits BGP features
that are expected to improve scalability (for instance, providing a
means to offload some of the processing burden associated with
client multicast routing onto one or many BGP route-reflectors).
This advantage has a cost (the maintenance of a amount of state
linear with the number of PEs joined to a stream), but when route
reflectors are used, this cost is spread among the route
reflectors.</t>
<t>However, the fourth mechanism is specific in that it offers the
possibility of offloading customer multicast routing processing onto
one or more BGP Route Reflector(s). When this is used, there is a
drawback of increasing the processing load placed on the route
reflector infrastructure. In the higher scale scenarios, it may be
required to adapt the route reflector infrastructure to the mVPN
routing load by using, for example:<list style="symbols">
<t>a separation of resources for unicast and multicast VPN
routing : using dedicated mVPN Route Reflector(s) (or using
dedicated mVPN BGP sessions or dedicated mVPN BGP instances)
;</t>
<t>the deployment of additional route reflector resources, for
example increasing the processing resources on existing route
reflectors or deployment of additional route reflectors.</t>
</list>Among the above, the most straightforward approach is to
consider the introduction of route reflectors dedicated to the mVPN
service and dimension them accordingly to the need of that service
(but doing so is not required and is left as an operator engineering
decision).</t>
</section>
<section title="P-routers scalability">
<t>Mechanisms (1) and (2) are restricted to use within multicast
VPNs that use an MI-PMSI, thereby necessitating:<list style="hanging">
<t hangText="">the use of a P-multicast tree technique that
allows shared trees (for example PIM-SM in ASM mode or MP2MP
LDP)</t>
<t hangText="or ">the use of one P-multicast tree per PE per
VPN, even for PEs that do not have sources in their directly
attached sites for that VPN.</t>
</list>By comparison, the fourth mechanism doesn't impose either
of these restrictions, and when P2MP trees are used only
necessitates the use of one tree per VPN per PE attached to a site
with a multicast source or RP (or with a candidate BSR, if BSR is
used).</t>
<t>In cases where there are less PEs connected with sources than the
total amount of PEs, it improves the amount of state maintained by
P-routers compared to the amount required to build an MI-PMSI with
P2MP trees. Such cases are expected to be frequent for multicast VPN
deployments (see <xref target="RFC4834">sections 4.2.4.1
of</xref>).</t>
</section>
<section title="Impact of C-multicast routing on Inter-AS deployments">
<t>Furthermore, co-existence with unicast inter-AS VPN options, and
an equal level of security for multicast and unicast including in an
inter-AS context, are specifically mentioned in <xref target="RFC4834">sections 5.2.6, 5.2.8 and 5.2.12 of</xref>.</t>
<t>In an inter-AS option B context, an isolation of ASes is obtained
as PEs don't have visibility of, nor exchange with, PEs of other
ASes. This property can be preserved if the segmented inter-AS
approach and BGP-based C-multicast routing is used, but it is not
preserved if PIM-based signaling is used.</t>
<t>By comparison, the fourth option (the use of BGP for carrying
C-Multicast routing) does not have any of the above limitations
related to inter-AS deployments.</t>
<t>Additionally, the authors note that the proposed BGP-based
approach for C-multicast routing provides a good fit with both the
segmented and non-segmented inter-AS approaches. By contrast, though
the PIM-based C-multicast routing is usable with segmented inter-AS
trees, the inter-AS scalability advantage of the approach is lost,
since PEs in an AS will see the C-multicast routing activity of all
other PEs of all other ASes.</t>
</section>
<section title="Security and robustness">
<t>BGP supports MD5 authentication of its peers for additional
security, thereby possibly benefit directly to multicast VPN
customer multicast routing, whether for intra-AS or inter-AS
communications. By contrast, with a PIM-based approach, no mechanism
providing a comparable level of security to authenticate
communications between remote PEs has been yet fully described yet
<xref target="I-D.ietf-pim-sm-linklocal"/>[], and in any case
would require significant additional operations for the provider to
be usable in a multicast VPN context.</t>
<t>The robustness of the infrastructure, especially the existing
infrastructure providing unicast VPN connectivity, is key. The
C-multicast routing function, especially under load, will compete
with the unicast routing infrastructure. With the PIM-based
approaches, the unicast and multicast VPN routing functions are
expected to only compete in the PE, for control plane processing
resources. In the case of the BGP-based approach, they will compete
on the PE for processing resources, and in the route reflectors
(supposing they are used for mVPN routing). It is identified that in
both cases, mechanisms will be required to arbitrate resources (e.g.
processing priorities). In the case of PIM-based procedures, between
the different control plane routing instances in the PE. And in the
case of the BGP-based approach, this is likely to require using
distinct BGP sessions for multicast and unicast (e.g. through the
use of dedicated mVPN BGP route reflectors, or to the use of a
distinct session with an existing route reflector).</t>
<t>Multicast routing is dynamic by nature, and multicast VPN routing
has to follow the VPN customers multicast routing events. The
different approaches can be compared on how they are expected to
behave in scenarios where multicast routing in the VPNs is subject
to an intense activity. Scalability of each approach under such a
load is detailed in <xref target="PEPE-mrouting-load"/>, and
the fourth approach (BGP-based) is the only one having a O(1) cost
for join/leave operations, and with which state maintenance is not
concentrated on the upstream PE.</t>
<t>On the other hand, while the BGP-based approach is likely to
suffer a slowdown under a load that is greater than the available
processing resources (because of possibly congested TCP sockets),
the PIM-based approaches would react to such a load by dropping
messages, with failure-recovery obtained through message refreshes.
Thus, the BGP-based approach could result in a degradation of
join/leave latency performance typically spread evenly across all
multicast streams being joined in that period, while the PIM-based
approach could result in increased join/leave latency, for some
random streams, by a multiple of the time between refreshes (e.g.
tens of seconds), and possibly in some states the adjacency may
time-out resulting in disruption of multicast streams.</t>
<t>The behavior of the PIM-based approach under such a load is also
harder to predict, given that the performance of the "Join
suppression" mechanism (an important mechanism for this approach to
scale) will itself be impeded by delays in Join processing. For
these reasons, the BGP-based approach would be able to provide a
smoother degradation and more predictable behavior under a highly
dynamic load.</t>
<t>In fact, both an "evenly spread degradation" and an "unevenly
spread larger degradation" can be problematic, and what seems
important is the ability for the VPN backbone operator to (a) limit
the amount of multicast routing activity that can be triggered by a
multicast VPN customer, and to (b) provide the best possible
independence between distinct VPNs. It seems that both of these can
be addressed through local implementation improvements, and that
both the BGP-based and PIM-based approaches could be engineered to
provide (a) and (b). It can be noted though that the BGP approach
proposes ways to dampen C-multicast route withdrawals and/or
advertisements, and thus already describes a way to provide (a),
while nothing comparable has yet been described for the PIM-based
approaches (even though it doesn't appear difficult). The PIM-based
approaches rely on a per VPN dataplane to carry the mVPN control
plane, and thus may benefit from this first level of separation to
solve (b).</t>
</section>
<section title="C-multicast VPN join latency">
<t><xref target="RFC4834">Section 5.1.3 of</xref> states that "the
group join delay [...] is also considered one important QoS
parameter. It is thus RECOMMENDED that a multicast VPN solution be
designed appropriately in this regard". In a multicast VPN context,
the "group join delay"of interest is the time between a CE sending a
PIM Join to its PE and the first packet of the corresponding
multicast stream being received by the CE.</t>
<t>It is to be noted that the C-multicast routing procedures will
only impact the group join latency of a said multicast stream for
the first receiver that is located across the provider backbone from
the multicast source-connected PE (or the first <n> receivers
in the specific case where a specific UMH selection algorithm is
used, that allows <n> distinct UMH to be selected by distinct
downstream PEs).</t>
<t>The different approaches proposed seem to have different
characteristics in how they are expected to impact join
latency:<list style="symbols">
<t>the PIM-based approaches minimize the number of control plane
processing hops between a new receiver-connected PE and the
source-connected PE, and being datagram-based introduces minimal
delay, thereby possibly having a join latency as good as
possible depending on implementation efficiency</t>
<t>under degraded conditions (packet loss, congestion, high
control plane load) the PIM-based approach may impact the
latency for a given multicast stream in an all or nothing manner
: if a C-multicast routing PIM Join packet is lost, latency can
reach a high time (a multiple of the periodicity of PIM Join
refreshes)</t>
<t>the BGP-based approach uses TCP exchanges, that may introduce
an additional delay depending on BGP and TCP implementation, but
which would typically result, under degraded conditions (such
packet loss, congestion, high control plane load), in a
comparably lower increase of latency spread more evenly across
the streams</t>
<t>as shown in <xref target="PEPE-mrouting-load"/>, the
BGP-based approach is particular in that it removes load from
all the PEs (without putting this load on the upstream PE for a
stream); this improvement of background load can bring improved
performance when a PE acts as the upstream PE for a stream, and
thus benefit join latency</t>
</list></t>
<t>This qualitative comparison of approaches shows that the
BGP-based approach is designed for a smoother degradation of latency
under degraded conditions such as packet loss, congestion, or high
control plane load. On the other hand, the PIM-based approaches seem
to structurally be able to reach the shorter "best-case" group join
latency (especially compared to deployment of the BGP-based approach
where route-reflectors are used).</t>
<t>Doing a quantitative comparison of latencies is not possible
without referring to specific implementations and benchmarking
procedures, and would possibly expose different conclusions,
especially for best-case group join latency for which performance is
expected vary with PIM and BGP implementations. We can also note
that improving a BGP implementation for reduced latency of route
processing would not only benefit multicast VPN group join latency,
but the whole BGP-based routing, which means that the need for good
BGP/RR performance is not specific to multicast VPN routing.</t>
<t>Last, C-multicast join latency will be impacted by the overall
load put on the control plane, and the scalability of the
C-multicast routing approach is thus to be taken into account. As
explained in sections <xref target="pepe-scaling"/> and <xref target="PEPE-mrouting-load"/>, the BGP-based approach will
provide the best scalability with an increased number of PEs per
VPN, thereby benefiting group join latency in such higher scale
scenarios.</t>
</section>
<section title="Extranet">
<t>An illustrative example of the benefit brought by using a
C-multicast routing approach close to the technique for unicast VPN
routing is how the "extranet" feature can be implemented : when
BGP-based mechanisms are used, the already defined and well
understood BGP route target import/export semantics are just reused
and applied to BGP mVPN routes. By contrast, it is not specified how
implementing the same feature would be done in the context of other
C-multicast routing mechanisms, and thus unclear how this would
bring a comparable consistency benefit, or if it is possible without
significant engineering trade-offs given that their control plane is
tied to a specific MI-PMSI tunnel. [to be updated when Extranet is
described for approaches other than the BGP-based approaches]</t>
<t>Note that the support for the Extranet feature is stated as a
MUST in <xref target="RFC4834">sections 5.1.6 of</xref>.</t>
</section>
<section title="Conclusion on C-multicast routing">
<t>The first and fourth approaches are relevant contenders for
C-multicast routing. Comparisons from a theoretical standpoint lead
to identify some advantages in the fourth approach, but possible
drawbacks are also identified for this approach. Comparisons from a
practical standpoint are harder to make, since only reduced
deployment and implementation information is available for the
fourth approach, but by default advantages would be seen in the
first approach has been applied through multiple deployments and
shown to be operationally viable.</t>
<t>Moreover, the first mechanism (full per-MVPN PIM peering across
an MI-PMSI) is the mechanism used by <xref target="I-D.rosen-vpn-mcast"/> and therefore it is deployed
and operating in MVPNs today. The fourth approach may or may not end
up being preferred for a said deployment, but because the first
approach has been in deployment for some time, the support for this
mechanism will in any case be helpful for to facilitate an eventual
migration from a deployment using mechanism close to the first
approach.</t>
<t>Consequently, at the present time, implementations are
recommended to support both the fourth (BGP-based) and first (Full
per-MPVN PIM peering) mechanisms. Further experience on deployments
of the fourth approach is needed before some best practice can be
defined.</t>
</section>
</section>
<section title="Encapsulation techniques for P-multicast trees">
<t>In this section the authors will not make any restricting
recommendations since the appropriateness of a specific provider core
data plane technology will depend on a large number of factors, for
example the service provider's currently deployed unicast data plane,
many of which are service provider specific.</t>
<t>However, implementations should not unreasonably restrict the data
plane technology that can be used, and should not force the use of the
same technology for different VPNs attached to a single PE. Initial
implementations may only support a reduced set of encapsulation
techniques and data plane technologies but this should not be a
limiting factor that hinders future support for other encapsulation
techniques, data plane technologies or interoperability.</t>
<t><xref target="RFC4834">Section 5.2.4.1 of</xref> states "In a
multicast VPN solution extending a unicast L3 PPVPN solution,
consistency in the tunneling technology has to be favored: such a
solution SHOULD allow the use of the same tunneling technology for
multicast as for unicast. Deployment consistency, ease of operation
and potential migrations are the main motivations behind this
requirement."</t>
<t>Current unicast VPN deployments use a variety of LDP, RSVP-TE and
GRE/IP-Multicast for encapsulating customer packets for transport
across the provider core of VPN services. In order to allow the same
encapsulations to be used for unicast and multicast VPN traffic, it is
recommended that multicast VPN standards should recommend
implementations to support for multicast VPNs, all the P2MP variants
of the encapsulations and signaling protocols that they support for
unicast and for which some multipoint extension is defined, such as
mLDP, P2MP RSVP-TE and GRE/IP-multicast.</t>
<t>All three of the above encapsulation techniques support the
building of P2MP multicast trees. In addition mLDP and
GRE/IP-ASM-Multicast implementations may also support the building of
MP2MP multicast trees. The use of MP2MP trees may provide some scaling
benefits to the service provider as only a single MP2MP tree need be
deployed per VPN, thus reducing by an order of magnitude the amount of
multicast state that needs to be maintained by P routers. This gain in
state is at the expense of bandwidth optimization, since sites that do
not have multicast receivers for multicast streams sourced behind a
said PE group will still receive packets of such streams, leading to
non-optimal bandwidth utilization across the VPN core. One thing to
consider is that the use of MP2MP multicast tree will require
additional configuration to define the same tree identifier or
multicast ASM group address in all PEs (it has been noted that some
auto-configuration could be possible for MP2MP trees, but this it is
not currently supported by the auto-discovery procedures). [ It has
been noted that C-multicast routing schemes not covered in <xref target="I-D.ietf-l3vpn-2547bis-mcast"/> could expose different
advantages of MP2MP multicast trees - this is out of scope of this
document ]</t>
<t>MVPN services can also be supported over a unicast VPN core through
the use of ingress PE replication whereby the ingress PE replicates
any multicast traffic over the P2P tunnels used to support unicast
traffic. While this option does not require the service provider to
modify their existing P routers (in terms of protocol support) and
does not require maintaining multicast-specific state on the P routers
in order for the service provider to be able deploy a multicast VPN
service, the use of ingress PE replication obviously leads to
non-optimal bandwidth utilization and it is therefore unlikely to be
the long term solution chosen by service providers. However ingress PE
replication may be useful during some migration scenarios or where a
service provider considers the level of multicast traffic on their
network to be too low to justify deploying multicast specific support
within their VPN core.</t>
<t>All proposed approaches for control plane and dataplane can be used
to provide aggregation amongst multicast groups within a VPN and
amongst different multicast VPNs, and potentially reduce the amount of
state to be maintained by P routers. However the latter -- the
aggregation amongst different multicast VPNs will require support for
upstream-assigned labels on the PEs. Support for upstream-assigned
labels may require changes to the data plane processing of the PEs and
this should be taken into consideration by service providers
considering the use of aggregate S-PMSI tunnels for the specific
platforms that the service provider has deployed.</t>
</section>
<section anchor="interas" title="Inter-AS deployments options">
<t>There are a number of scenarios that lead to the requirement for
inter-AS multicast VPNs, including:<list style="numbers">
<t>a service provider may have a large network that they have
segmented into a number of ASs.</t>
<t>a service provider's multicast VPN may consist of a number of
ASs due to acquisitions and mergers with other service
providers.</t>
<t>a service provider may wish to interconnect their multicast VPN
platform with that of another service provider.</t>
</list>The first scenario can be considered the "simplest" because
the network is wholly managed by a single service provider under a
single strategy and is therefore likely to use a consistent set of
technologies across each AS.</t>
<t>The second scenario may be more complex than the first because the
strategy and technology choices made for each AS may have been
different due to their differing history and the service provider may
not have (or may be unwilling to) unified the strategy and technology
choices for each AS.</t>
<t>The third scenario is the most complex because in addition to the
complexity of the second scenario, the ASs are managed by different
service providers and therefore may be subject to a different trust
model than the other scenarios.</t>
<t><xref target="RFC4834">Section 5.2.6 of</xref> states that "a
solution MUST support inter-AS multicast VPNs, and SHOULD support
inter-provider multicast VPNs", "considerations about coexistence with
unicast inter-AS VPN Options A, B and C (as described in section 10 of
[RFC4364]) are strongly encouraged" and "a multicast VPN solution
SHOULD provide inter-AS mechanisms requiring the least possible
coordination between providers, and keep the need for detailed
knowledge of providers' networks to a minimum - all this being in
comparison with corresponding unicast VPN options".</t>
<t><xref target="I-D.ietf-l3vpn-2547bis-mcast">Section 8 of </xref>
addresses these requirements by proposing two approaches for mVPN
inter-AS deployments:</t>
<t><list style="numbers">
<t>Non-segmented inter-AS tunnels where the multicast tunnels are
end-to-end across ASes, so even though the PEs belonging to a
given MVPN may be in different ASs the ASBRs play no special role
and function merely as P routers (described in section 8.1).</t>
<t>Segmented inter-AS tunnels where each AS constructs its own
separate multicast tunnels which are then 'stitched' together by
the ASBRs (described in section 8.2).</t>
</list></t>
<t><xref target="RFC4834">Section 5.2.6 of</xref> also states "Within
each service provider the service provider SHOULD be able on its own
to pick the most appropriate tunneling mechanism to carry (multicast)
traffic among PEs (just like what is done today for unicast)". The
segmented approach is the only one capable of meeting this
requirement.</t>
<t>The segmented inter-AS solution would appear to offer the largest
degree of deployment flexibility to operators. However the
non-segmented inter-AS solution can simplify deployment in a
restricted number of scenarios and <xref target="I-D.rosen-vpn-mcast"/> only supports the non-segmented
inter-AS solution and therefore the non-segmented inter-AS solution is
likely to be useful to some operators for backward compatibility and
during migration from <xref target="I-D.rosen-vpn-mcast"/> to
<xref target="I-D.ietf-l3vpn-2547bis-mcast"/>.</t>
<t>The applicability of segmented or non-segmented inter-AS tunnels to
a given deployment or inter-provider interconnect will depend on a
number of factors specific to each service provider. However, due to
the additional deployment flexibility offered by segmented inter-AS
tunnels, it is the recommendation of the authors that all
implementations should support the segmented inter-AS model.
Additionally, the authors recommend that implementations should
consider supporting the non-segmented inter-AS model in order to
facilitate co-existence with existing deployments, and as a feature to
provide a lighter engineering in a restricted set of scenarios,
although it is recognized that initial implementations may only
support one or the other.</t>
</section>
<section anchor="bidir" title="Bidir-PIM support">
<t>In Bidir-PIM, the packet forwarding rules have been improved over
PIM-SM, allowing traffic to be passed up the shared tree toward the RP
Address (RPA). To avoid multicast packet looping, Bidir-PIM uses a
mechanism called the designated forwarder (DF) election, which
establishes a loop-free tree rooted at the RPA. Use of this method
ensures that only one copy of every packet will be sent to an RPA,
even if there are parallel equal cost paths to the RPA. To avoid loops
the DF election process enforces consistent view of the DF on all
routers on network segment, and during periods of ambiguity or routing
convergence the traffic forwarding is suspended.</t>
<t>In the context of a multicast VPN solution, a solution for
Bidir-PIM support must preserve this property of similarly avoiding
packet loops, including in the case where mVRF's in a given MVPN don't
have a consistent view of the routing to C-RPL/C-RPA.</t>
<t>The current MVPN specifications <xref target="I-D.ietf-l3vpn-2547bis-mcast"/> in section 11, define
three methods to support Bidir-PIM, as RECOMMENDED in <xref target="RFC4834"/> :<list style="numbers">
<t>Standard DF election procedure over an MI-PMSI</t>
<t>VPN Backbone as the RPL (section 11.1)</t>
<t>Partitioned Sets of PEs (section 11.2)</t>
</list></t>
<t>Method (1) is naturally applied to deployments using "Full per-MVPN
PIM peering across an MI-PMSI" for C-multicast routing, but as
indicated in <xref target="I-D.ietf-l3vpn-2547bis-mcast"/> in
section 11, the DF Election may not work well in an mVPN environment
and an alternative to DF election would be desirable.</t>
<t>The advantage of method (2) and (3) is that they do not require
running the DF election procedure among PEs.</t>
<t>Method (2) leverages the fact that in Bidir-PIM, running the DF
election procedure is not needed on the RPL. This approach thus has
the benefit of simplicity of implementation, especially in a context
where BGP-based C-multicast routing is used. However it has the
drawback of putting constraints on how Bidir-PIM is deployed which may
not always match mVPN customers requirements.</t>
<t>Method (3) treats an mVPN as a collection of sets of multicast
VRFs, all PEs in a set having the same reachability information
towards C-RPA, but distinct from PEs in other sets. Hence, with this
method, C-Bidir packet loops in MVPN are resolved by the ability to
partition a VPN into disjoints sets of VRF's, each having a distinct
view of converged network. The partitioning approach to Bidir-PIM
requires either upstream-assigned MPLS labels (to denote the
partition) or a unique MP2MP LSP per partition. The former is based on
PE Distinguisher Labels that have to be distributed using
auto-discovery BGP routes and their handling requires the support for
upstream assigned labels and context label lookups [ref]. The latter,
using MP2MP LSP per partition, does not have these constraints but is
restricted to P-tunnel types supporting MP2MP connectivity (such as
mLDP[ref]).</t>
<t>This approach to C-Bidir can work with PIM-based or BGP-based
C-multicast routing procedures, and is also generic in the sense that
it does not impose any requirements on the Bidir-PIM service
offering.</t>
<t>Given the above considerations, method (3) "Partitioned Sets of
PEs" is the RECOMMENDED approach.</t>
<t>In the event where method (3) is not applicable (lack of support
for upstream assigned labels or for a P-tunnel type providing MP2MP
connectivity), then method (1) "Standard DF election procedure over an
MI-PMSI" and (2) "VPN Backbone as the RPL" are RECOMMENDED as interim
solutions, (1) having the advantage over (2) of not putting
constraints on how Bidir-PIM is deployed and the drawbacks of only
being applicable when PIM-based C-multicast is used and of possibly
not working well in an mVPN environment.</t>
</section>
</section>
<section title="Co-located RPs">
<t><xref target="RFC4834">Section 5.1.10.1 of</xref> states "In the case
of PIM-SM in ASM mode, engineering of the RP function requires the
deployment of specific protocols and associated configurations. A
service provider may offer to manage customers' multicast protocol
operation on their behalf. This implies that it is necessary to consider
cases where a customer's RPs are out-sourced (e.g. on PEs).
Consequently, a VPN solution MAY support the hosting of the RP function
in a VR or VRF."</t>
<t>However, customers who have already deployed multicast within their
networks and have therefore already deployed their own internal RPs are
often reluctant to hand over the control of their RPs to their service
provider and make use of a co-located RP model, and providing
RP-collocation on a PE will require the activation of MSDP or the
processing of PIM Registers on the PE. Securing the PE routers for such
activity requires special care, additional work, and will likely rely on
specific features to be provided by the routers themselves.</t>
<t>The applicability of the co-located RP model to a given MVPN will
thus depend on a number of factors specific to each customer and service
provider.</t>
<t>It is therefore the recommendation that implementations should
support a co-located RP model, but that support for a co-located RP
model within an implementation should not restrict deployments to using
a co-located RP model : implementations MUST support deployments when
activation of a PIM RP function (PIM Register processing and RP-specific
PIM procedures) or VRF MSDP instance is not required on any PE router
and where all the RPs are deployed within the customers' networks or
CEs.</t>
</section>
<section title="Existing deployments">
<t>Some suggestions provided in this document can be used to
incrementally modify currently deployed implementations without
hindering these deployments, and without hindering the consistency of
the standardized solution by providing optional per-VRF configuration
knobs to support modes of operation compatible with currently deployed
implementations, while at the same time using the recommended approach
on implementations supporting the standard.</t>
<t>In cases where this may not be easily achieved, a recommended
approach would be to provide a per-VRF configuration knob that allows
incremental per-VPN migration of the mechanisms used by a PE device,
which would allow migration with some per-VPN interruption of service
(e.g. during a maintenance window).</t>
<t>Mechanisms allowing "live" migration by providing concurrent use of
multiple alternatives for a given PE and a given VPN, is not seen as a
priority considering the expected implementation complexity associated
with such mechanisms. However, if there happen to be cases where they
could be viably implemented relatively simply, such mechanisms may help
improve migration management.</t>
</section>
<section anchor="summary" title="Summary of recommendations">
<t>The following list summarizes conclusions on the mechanisms that
define the set of mandatory to implement mechanisms in the context of
<xref target="I-D.ietf-l3vpn-2547bis-mcast"/>.</t>
<t>Note well that the implementation of the non-mandatory alternative
mechanisms is not precluded.</t>
<t>Recommendations are:<list style="symbols">
<t>that BGP-based auto-discovery be the mandated solution for
auto-discovery ;</t>
<t>that BGP be the mandated solution for S-PMSI switching signaling
;</t>
<t>that implementations support both the BGP-based and the full
per-MPVN PIM peering solutions for PE-PE transmission of customer
multicast routing until further operational experience is gained
with both solutions ;</t>
<t>that implementations use the "Partitioned Sets of PEs" approach
for Bidir-PIM support ;</t>
<t>that implementations implement the P2MP variants of the P2P
protocols that they already implement, such as mLDP, P2MP RSVP-TE
and GRE/IP-Multicast ;</t>
<t>that implementations support segmented inter-AS tunnels and
consider supporting non-segmented inter-AS tunnels (in order to
maintain backwards compatibility and for migration) ;</t>
<t>implementations MUST support deployments when activation of a PIM
RP function (PIM Register processing and RP-specific PIM procedures)
or VRF MSDP instance is not required on any PE router.</t>
</list></t>
</section>
<section anchor="IANA" title="IANA Considerations">
<t>This document makes no request to IANA.</t>
<t>[ Note to RFC Editor: this section may be removed on publication as
an RFC. ]</t>
</section>
<section anchor="Security" title="Security Considerations">
<t>This document does not by itself raise any particular security
considerations.</t>
</section>
<section anchor="Acknowledgements" title="Acknowledgements">
<t>We would like to thank Adrian Farrel, Eric Rosen, Yakov Rekhter, and
Maria Napierala for their feedback that helped shape this document.</t>
<t>Additional credit is due to Maria Napierala for co-authoring <xref target="bidir"/> on <xref format="title" target="bidir"/>.</t>
</section>
</middle>
<back>
<references title="Informative References">
<reference anchor="RFC4834">
<front>
<title>Requirements for Multicast in L3 Provider-Provisioned Virtual
Private Networks (PPVPNs)</title>
<author fullname="Thomas Morin" initials="T" surname="Morin">
<organization/>
</author>
<date day="" month="April" year="2007"/>
<abstract>
<t>This document presents a set of functional requirements for
network solutions that allow the deployment of IP multicast within
L3 Provider Provisioned Virtual Private Networks (PPVPNs). It
specifies requirements both from the end user and service provider
standpoints. It is intended that potential solutions specifying
the support of IP multicast within such VPNs will use these
requirements as guidelines.</t>
</abstract>
</front>
<seriesInfo name="RFC" value="4834"/>
<format target="http://www.ietf.org/rfc/rfc4834.txt" type="TXT"/>
</reference>
<?rfc include='reference.I-D.ietf-l3vpn-2547bis-mcast'?>
<?rfc include='reference.I-D.ietf-l3vpn-2547bis-mcast-bgp'?>
<?rfc include='reference.I-D.rosen-vpn-mcast'?>
<reference anchor="I-D.raggarwa-l3vpn-2547-mvpn">
<front>
<title>Base Specification for Multicast in BGP/MPLS VPNs</title>
<author fullname="Rahul Aggarwal" initials="R" surname="Aggarwal">
<organization/>
</author>
<date day="22" month="June" year="2004"/>
<abstract>
<t>This document describes the minimal set of procedures required
to build multi-vendor inter-operable implementations of multicast
for BGP/MPLS VPNs. It is based on prior specifications of
multicast for BGP/MPLS VPN specifications that have been
implemented and deployed. The procedures described herein require
PIM-SM as the multicast routing protocol in the SP network.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-raggarwa-l3vpn-2547-mvpn-00"/>
<format target="http://www.ietf.org/internet-drafts/draft-raggarwa-l3vpn-2547-mvpn-00.txt" type="TXT"/>
</reference>
<reference anchor="I-D.ietf-pim-sm-linklocal">
<front>
<title>Authentication and Confidentiality in PIM-SM Link-local
Messages</title>
<author fullname="John Atwood" initials="J" surname="Atwood">
<organization/>
</author>
<date day="18" month="November" year="2007"/>
<abstract>
<t>RFC 4601 mandates the use of IPsec to ensure authentication of
the link-local messages in the Protocol Independent Multicast -
Sparse Mode (PIM-SM) routing protocol. This document specifies
mechanisms to authenticate the PIM-SM link local messages using
the IP security (IPsec) Authentication Header (AH) or
Encapsulating Security Payload (ESP). It specifies optional
mechanisms to provide confidentiality using the ESP. Manual keying
is specified as the mandatory and default group key management
solution. To deal with issues of scalability and security that
exist with manual keying, an optional automated group key
management mechanism is specified.</t>
</abstract>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-pim-sm-linklocal-02"/>
<format target="http://www.ietf.org/internet-drafts/draft-ietf-pim-sm-linklocal-02.txt" type="TXT"/>
</reference>
<?rfc include='reference.I-D.ietf-pim-port'?>
<?rfc include='reference.RFC.4684'?>
<reference anchor="RFC2119">
<front>
<title abbrev="RFC Key Words">Key words for use in RFCs to Indicate
Requirement Levels</title>
<author fullname="Scott Bradner" initials="S." surname="Bradner">
<organization>Harvard University</organization>
<address>
<postal>
<street>1350 Mass. Ave.</street>
<street>Cambridge</street>
<street>MA 02138</street>
</postal>
<phone>- +1 617 495 3864</phone>
<email>sob@harvard.edu</email>
</address>
</author>
<date month="March" year="1997"/>
<area>General</area>
<keyword>keyword</keyword>
<abstract>
<t>In many standards track documents several words are used to
signify the requirements in the specification. These words are
often capitalized. This document defines these words as they
should be interpreted in IETF documents. Authors who follow these
guidelines should incorporate this phrase near the beginning of
their document: <list>
<t>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 RFC 2119.</t>
</list></t>
<t>Note that the force of these words is modified by the
requirement level of the document in which they are used.</t>
</abstract>
</front>
<seriesInfo name="BCP" value="14"/>
<seriesInfo name="RFC" value="2119"/>
<format octets="4723" target="ftp://ftp.isi.edu/in-notes/rfc2119.txt" type="TXT"/>
<format octets="16553" target="http://xml.resource.org/public/rfc/html/rfc2119.html" type="HTML"/>
<format octets="5703" target="http://xml.resource.org/public/rfc/xml/rfc2119.xml" type="XML"/>
</reference>
</references>
<section anchor="PEPE-mrouting-load" title="Scalability of C-multicast routing processing load">
<t>The main role of multicast routing is to let routers determine that
they should start or stop forwarding a said multicast stream on a said
link. In the multicast VPN context, this has to be done for each VPN,
and the associated function is thus named "customer-multicast routing"
or "C-multicast routing" and its role is to let PE routers determine
that they should start or stop forwarding the traffic of a said
multicast stream toward the remote PEs, on some PMSI tunnel.</t>
<t>When some "join" message is received by a PE, this PE knows that it
should be sending traffic for the corresponding multicast group of the
corresponding VPN. But the reception of a "prune" message from a remote
PE is not enough by itself for a PE to know that it should stop
forwarding the corresponding multicast traffic : it has to make sure
that they aren't any other PEs that still have receivers for this
traffic.</t>
<t>There are many ways that the "C-multicast routing" building block can
be designed, and they differ, among other things, in how a PE determines
when it can stop forwarding a said multicast stream toward other
PEs:<list style="hanging">
<t hangText="PIM LAN Procedures, by default"><vspace blankLines="0"/>By default when PIM LAN procedures are used, when a
PE Prunes itself from a multicast tree, all other PEs check their
own state to known if they are on the tree, in which case they send
a PIM Join message to override the Prune. The "did the last receiver
leave?" question is thus implicitly answered by all PE routers, for
each PIM Prune message.</t>
<t hangText="PIM LAN Procedures, with explicit tracking : "><vspace blankLines="0"/>PIM LAN procedures can use an "explicit tracking"
approach, where a PE which is the upstream router for a multicast
stream maintains an updated list of all neighbors who are joined to
the tree. Thus, when it receives a Leave message from a PIM
neighbor, it instantly knows the answer to the "did the last
receiver leave?" question.<vspace blankLines="0"/>In this case, the
question is answered by the upstream router alone. The side effect
of this "explicit tracking" is that "Join suppression" is not used :
the downstream PEs will always send Joins toward the upstream PE,
which will have to process them all.</t>
<t hangText="BGP-based C-multicast routing"><vspace blankLines="0"/>When BGP-based procedures are used for C-multicast
routing, if no BGP route reflector is used, the "did the last
receiver leave?" question is answered like in the PIM "explicit
tracking" approach.<vspace blankLines="0"/>But, when a BGP route
reflector is used (which is expected to be the recommended
approach), the role of maintaining an updated list of the PE part of
a said multicast tree is taken care of by the route reflector(s).
Using plain BGP route selection procedures, the route reflector will
withdraw a C-multicast Source Tree Join for a said (C-S,C-G) when
there is no PE advertising one anymore. In this context, the "did
the last receiver leave?" question can be said to be answered by the
route-reflector(s).<vspace blankLines="0"/>Furthermore, the BGP
route distribution can leverage more than one route reflector : if a
hierarchy of route reflectors is used, the "did the last receiver
leave?" question is partly answered by each route reflector in the
hierarchy.</t>
</list></t>
<t>We can see that answering the "last receiver leaves" question is a
significant proportion of the work that the C-multicast routing building
block has to make, and where approaches differ most. The different
approaches for handling C-multicast routing can result in a different
amount of processing and how this processing is spread among the
different functions. These differences can be better estimated by
quantifying the amount of message processing and state maintenance.</t>
<t>Though the type of processing, messages and states, may vary with the
different approaches, we propose here a rough estimation of the load of
PEs, in terms of number of messages processed and number of control
plane states maintained : a "message processed" being a message being
parsed, a lookup being done, and some action being taken (such has
updating a control plane or data plane state), and a "state maintained"
being a multicast state kept in the control plane memory of a PE,
related to a interface or a PE being subscribed to a multicast stream
(we don't compare the data plane states on PE routers, which wouldn't
vary between the different options chosen).</t>
<t>The following subsections do such an estimation for each proposed
approach for C-multicast routing, for different phases of the following
scenario:</t>
<t><list style="symbols">
<t>one SSM multicast stream is considered (extrapolating to a higher
number of streams is linear)</t>
<t>only the intra-AS case is concerned (with the segmented inter-AS
trees and BGP-based C-multicast routing, #mvpn_PE and #R_PE should
refer to the PEs of the mVPN in the AS, not to all PEs of the
mVPN)</t>
<t>the scenario is as follows:<list style="symbols">
<t>one PE Joins the multicast stream (because of a new
receiver-connected site has sent a Join on the PE-CE link),
followed by a number of additional PEs that also join the
multicast stream, one after the other ; we evaluate the
processing required for the addition of each PE</t>
<t>some period of time T passes, without any PE joining or
leaving (baseline)</t>
<t>all PE leaves, one after the other, until the last one leaves
; we evaluate the processing required for the leave of each
PE</t>
</list></t>
<t>the parameters used are:<list style="symbols">
<t>#mVPN_PE : the number of PEs in the mVPN</t>
<t>#R_PE : the number of PEs joining the multicast stream</t>
<t>#RR : the number of route reflectors</t>
<t>T_PIM_r : the time between two refreshes of a PIM Join
(default is 60s)</t>
</list></t>
</list></t>
<t>The estimation unit used is the "message.equipment" or "m.e", one
"message.equipment" being "one equipment processing one message" (10 m.e
being "10 equipments processing each one message", or "5 messages each
processed by 2 equipments", or "1 message processed by 10 equipment",
etc.). Similarly, for the amount of control plane state, the unit used
is "state.equipment" or "s.e".</t>
<t>We distinguish three different types of equipments : the upstream PE
for the multicast stream, the RR (if any), and the other PEs (which are
not the upstream PE). The estimation is a total number of
"message.equipment", for each type of equipment.</t>
<t>Additional precisions:</t>
<t><list style="symbols">
<t>for PIM, only Join and Prune messages are counted ; the PIM
Hellos are not counted since these are not messages that trigger
specific action in a typical scenario; message processing related to
the PIM Assert mechanism is also not taken into account, because it
is only active in transient state</t>
<t>for BGP, only UPDATE messages for mVPN routes carrying
C-multicast routing information are considered</t>
</list></t>
<section title="PIM LAN procedures, by default">
<texttable style="all" title="Amount of messages processed for one multicast tree of one VPN - PIM LAN procedures, by default">
<ttcol width="12%"/>
<ttcol width="22%">upstream PE (1)</ttcol>
<ttcol width="22%">other PEs (#mvpn_PE-1)</ttcol>
<ttcol width="22%">RR (none)</ttcol>
<ttcol width="22%">total</ttcol>
<c>first PE joins</c>
<c>1 m.e</c>
<c>#mVPN_PE-1 m.e</c>
<c>/</c>
<c>#mVPN_PE m.e</c>
<c>for *each* additional PE joining</c>
<c>1 m.e</c>
<c>#mvpn_PE-1 m.e</c>
<c>/</c>
<c>#mvpn_PE m.e</c>
<c>baseline processing over a period T</c>
<c>T/T_PIM_r m.e</c>
<c>(T/T_PIM_r) . (#mvpn_PE-1) m.e</c>
<c>/</c>
<c>(T/T_PIM_r) x #mvpn_PE m.e</c>
<c>for *each* PE leaving</c>
<c>2 m.e</c>
<c>2(#mvpn_PE-1) m.e</c>
<c>/</c>
<c>2 x #mvpn_PE m.e</c>
<c>the last PE leaves</c>
<c>1 m.e</c>
<c>#mvpn_PE-1 m.e</c>
<c>/</c>
<c>#mvpn_PE m.e</c>
<c>total for #R_PE PEs</c>
<c>#R_PE x 2 + T/T_PIM_r m.e</c>
<c>(#mvpn_PE-1) x (#R_PE) x 2 + T/T_PIM_r) . (#mvpn_PE-1) m.e</c>
<c>0</c>
<c>#mvpn_PE x ( 3 x #R_PE + T/T_PIM_r ) m.e</c>
<c>total state maintained</c>
<c>1 s.e</c>
<c>#R_PE s.e</c>
<c>0</c>
<c>#R_PE+1 s.e</c>
</texttable>
<t>We suppose here that the PIM Join suppression and Prune Override
mechanisms are fully effective, i.e. that a Join or Prune message sent
by a PE is instantly seen by other PEs. Strictly speaking, this is not
true, and depending on network delays and timing, there could be cases
where more messages are exchanged and the number given in this table
is a lower bound to the number of PIM messages exchanged.</t>
</section>
<section anchor="pim-explicit" title="PIM LAN procedures, with explicit tracking">
<texttable style="all" title="Amount of messages processed for one multicast tree of one VPN - PIM LAN procedures, with explicit tracking">
<ttcol/>
<ttcol>upstream PE (1)</ttcol>
<ttcol>other PEs (#mvpn_PE-1)</ttcol>
<ttcol>RRs (none)</ttcol>
<ttcol>total</ttcol>
<c>first PE joins</c>
<c>1 m.e</c>
<c>1 m.e (see note below)</c>
<c>/</c>
<c>2 m.e</c>
<c>for *each* additional PE joining</c>
<c>1 m.e</c>
<c>1 m.e (see note below)</c>
<c>/</c>
<c>2 m.e</c>
<c>baseline processing over a period T</c>
<c>(T/T_PIM_r) m.e x #R_PE m.e</c>
<c>(T/T_PIM_r) m.e (see note below)</c>
<c>/</c>
<c>(T/T_PIM_r) x #R_PE m.e</c>
<c>for *each* PE leaving</c>
<c>1 m.e</c>
<c>1 m.e (see note below)</c>
<c>/</c>
<c>2 m.e</c>
<c>the last PE leaves</c>
<c>1 m.e</c>
<c>1 m.e (see note below)</c>
<c>/</c>
<c>2 m.e</c>
<c>total for #R_PE PEs</c>
<c>#R_PE (2 + T/T_PIM_r) m.e</c>
<c>#R_PE x ( 2 + T/T_PIM_r) m.e</c>
<c>0</c>
<c>#R_PE x ( 4 + T/T_PIM_r) m.e</c>
<c>total state maintained</c>
<c>#R_PE s.e</c>
<c>#R_PE s.e</c>
<c>0</c>
<c>2 x #R_PE s.e</c>
</texttable>
<t>Note: in this explicit tracking mode, a said Join or Leave message
requires processing only by the upstream PE and the PE sending the
message ; indeed, other PEs don't have any action to take ; it is to
be noted though that these other PEs will still have to parse the PIM
message, which is not zero processing. We make here the assumption
that this is not significant.</t>
</section>
<section title="BGP-based">
<t>About RR: we suppose that a message has to be processed by r BGP
route reflectors to go from a receiver-connected PE to the
source-connected PE. In practice, r depends on how RR are meshed, and
would typically be small (max 1,2,3...), and r tends quickly toward 1
(as soon as there is a receiver-connected PEs in each RR cluster).</t>
<t>We make the assumption that BGP constrained VPN route distribution
<xref target="RFC4684"/> is used, if not the amount of state and
message processing with this approach is similar to the PIM with
explicit tracking approach<xref target="pim-explicit"> </xref>,
without the Joins refreshes.</t>
<texttable style="all" title="Amount of messages processed for one multicast tree of one VPN - BGP-based procedures">
<ttcol/>
<ttcol>upstream PE (1)</ttcol>
<ttcol>other PEs (#mvpn_PE-1)</ttcol>
<ttcol>RRs (#RR)</ttcol>
<ttcol>total</ttcol>
<c>first PE joins</c>
<c>1 m.e</c>
<c>1 m.e</c>
<c>r m.e</c>
<c>(r+2) m.e</c>
<c>for *each* additional PE joining</c>
<c>0</c>
<c>1 m.e</c>
<c>between 1 and r m.e</c>
<c>between 2 and (r+1) m.e</c>
<c>baseline processing over a period T</c>
<c>0</c>
<c>0</c>
<c>0</c>
<c>0</c>
<c>for *each* PE leaving</c>
<c>0</c>
<c>1 m.e</c>
<c>between 1 and r m.e</c>
<c>between 2 and (r+1) m.e</c>
<c>the last PE leaves</c>
<c>1 m.e</c>
<c>1 m.e</c>
<c>r m.e</c>
<c>(r+2) m.e</c>
<c>total for #R_PE PEs</c>
<c>2 m.e</c>
<c>#R_PE x 2 m.e</c>
<c>2 (r+#R_PE) m.e</c>
<c>2 (2 x #R_PE + r + 1) m.e</c>
<c>total state maintained</c>
<c>2 s.e</c>
<c>2 x #R_PE s.e</c>
<c>approx. (#R_PE x #RR) s.e</c>
<c>approx. (#R_PE x (#RR+2)) m.e</c>
</texttable>
<t/>
</section>
<section anchor="quant-conclusion" title="Side by side orders of magnitude comparison">
<t>This section concludes on the previous section by considering the
orders of magnitude when the number of PEs in a VPN increases.</t>
<texttable style="all" title="Amount of messages processed for one multicast tree of one VPN - PIM LAN procedures, with explicit tracking">
<ttcol/>
<ttcol>PIM LAN Procedures, default</ttcol>
<ttcol>PIM LAN Procedures, explicit tracking</ttcol>
<ttcol>BGP-based</ttcol>
<c>first PE joins</c>
<c>O(#mVPN_PE)</c>
<c>O(1)</c>
<c>O(1)</c>
<c>for *each* additional PE joining</c>
<c>O(#mVPN_PE)</c>
<c>O(1)</c>
<c>O(1)</c>
<c>baseline processing over a period T</c>
<c>(T/T_PIM_r) x O(#mvpn_PE)</c>
<c>(T/T_PIM_r) x O(#R_PE)</c>
<c>0</c>
<c>for *each* PE leaving</c>
<c>O(#mVPN_PE)</c>
<c>O(1)</c>
<c>O(1)</c>
<c>the last PE leaves</c>
<c>O(#mVPN_PE)</c>
<c>O(1)</c>
<c>O(1)</c>
<c>total for #R_PE PEs</c>
<c>O(#mVPN_PE x #R_PE) + O(#mVPN_PE x T/T_PIM_r)</c>
<c>O(#R_PE) x (T/T_PIM_r)</c>
<c>O(#R_PE)</c>
<c>states</c>
<c>O(#R_PE)</c>
<c>O(#R_PE)</c>
<c>O(#R_PE x #RR)</c>
<c>notes</c>
<c>(processing and state maintenance are essentially done by, and
spread amongst, the PEs of the MVPN ; non-upstream PEs have
processing to do)</c>
<c>(processing and state maintenance is essentially done on the
upstream PE)</c>
<c>(processing and state maintenance is essentially done by, and
spread amongst, the RRs)</c>
</texttable>
<t>The conclusions that can be drawn from the above are that:</t>
<t><list style="symbols">
<t>the PIM LAN Procedures default approach is particular in that
all PEs, including those that are neither upstream nor downstream
for a given message have processing to do, which results in a
total amount of messages to process which is in O(#mVPN_PE x
#R_PE), i.e. O(#mVPN_PE ^ 2) if the proportion of receiver PEs is
considered constant when the number of PEs increases ;</t>
<t>the two PIM-based approach do refreshes of Join messages, this
is a linear factor not changing the order of magnitude, but which
can be significant for long-lived streams ;</t>
<t>the BGP-based approach requires an amount of message processing
in O(#R_PE), lower than the two other approaches, and which is
independent of the duration of streams ;</t>
<t>state maintenance is in the same order of magnitude for all
approaches : O(#R_PE), but the repartition is different:<list style="symbols">
<t>the PIM LAN Procedure default approach fully spreads, and
minimizes, the amount of state (one state per PE)</t>
<t>the PIM LAN procedure with explicit tracking, concentrate
all state on the upstream PE</t>
<t>the BGP-based procedures spread all the state on the set of
route reflectors</t>
</list></t>
</list></t>
<t>This quantification of message processing is based on a use case
where each PE with a receiver joins and leave once. Drawing
scalability-related conclusions for other patterns or frequency of
changes of the set of receiver-connected PEs, requires considering the
cost of each approach for "a new PE joining" and "a (non-last) PE
leaving". From this perspective, the "PIM LAN Procedure default
approach" is the most costly one (processing in O(#mVPN_PE)), whereas
the other approaches are in O(1) ; the "PIM LAN Procedures with
explicit tracking" reduce the processing to the minimum in that case,
the BGP-based approach having a cost increased by a linear factor
depending on the number of RRs that will have to parse the
message.</t>
</section>
</section>
<section title="Switching to S-PMSI">
<t>[ the following point was fixed in version 07 of <xref target="I-D.ietf-l3vpn-2547bis-mcast"/>, and is here for reference
only ]</t>
<t><xref target="I-D.ietf-l3vpn-2547bis-mcast">Section 7.2.2.3 of</xref>
proposes two approaches for how a source PE can decide when to start
transmitting customer multicast traffic on a S-PMSI:</t>
<t><list style="numbers">
<t>The source PE sends multicast packets for the <C-S, C-G> on
both the I-PMSI P-multicast tree and the S-PMSI P-multicast tree
simultaneously for a pre-configured period of time, letting the
receiver PEs select the new tree for reception, before switching to
only the S-PMSI.</t>
<t>The source PE waits for a pre-configured period of time after
advertising the <C-S, C-G> entry bound to the S-PMSI before
fully switching the traffic onto the S-PMSI-bound P-multicast
tree.</t>
</list>The first alternative has essentially two drawbacks:<list style="symbols">
<t><C-S,C-G> traffic is sent twice for some period of time,
which would appear to be at odds with the motivation for switching
to an S-PMSI in order to optimize the bandwidth used by the
multicast tree for that stream.</t>
<t>It is unlikely that the switchover can occur without packet loss
or duplication if the transit delays of the I-PMSI P-multicast tree
and the S-PMSI P-multicast tree differ.</t>
</list></t>
<t>By contrast, the second alternative has none of these drawbacks, and
satisfy the requirement in <xref target="RFC4834">section 5.1.3
of</xref>, which states that "[...] a multicast VPN solution SHOULD as
much as possible ensure that client multicast traffic packets are
neither lost nor duplicated, even when changes occur in the way a client
multicast data stream is carried over the provider network". The second
alternative also happen to be the one used in existing deployments.</t>
<t>For these reasons, it is the authors' recommendation to mandate the
implementation of the second alternative for switching to S-PMSI.</t>
</section>
</back>
</rfc>
| PAFTECH AB 2003-2026 | 2026-04-21 22:24:38 |