One document matched: draft-minto-2547-egress-node-fast-protection-03.xml


<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!ENTITY RFC2205 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2205.xml">
<!ENTITY RFC3031 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3031.xml">
<!ENTITY RFC3209 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3209.xml">
<!ENTITY RFC4090 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4090.xml">
<!ENTITY RFC5036 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5036.xml">
<!ENTITY RFC4447 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4447.xml">
<!ENTITY RFC3471 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3471.xml">
<!ENTITY RFC3472 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3472.xml">
<!ENTITY RFC5331 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5331.xml">
<!ENTITY RFC5920 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5920.xml">
]>
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<?rfc strict="yes" ?>
<?rfc toc="yes"?>
<?rfc tocdepth="4"?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="no" ?>
<?rfc compact="yes" ?>
<?rfc subcompact="no" ?>
<rfc category="std" docName="draft-minto-2547-egress-node-fast-protection-03"
     ipr="trust200902">
  <front>
    <title>2547 egress PE Fast Failure Protection</title>

    <author fullname="Jeyananth Minto Jeganathan" initials="J.M"
            surname="Jeganathan">
      <organization>Juniper Networks</organization>

      <address>
        <postal>
          <street>1194 N Mathilda Avenue</street>

          <city>Sunnyvale</city>

          <region>CA</region>

          <code>94089</code>

          <country>USA</country>
        </postal>

        <email>minto@juniper.net</email>
      </address>
    </author>

    <author fullname="Hannes Gredler" initials="H" surname="Gredler">
      <organization>Juniper Networks</organization>

      <address>
        <postal>
          <street>1194 N Mathilda Avenue</street>

          <city>Sunnyvale</city>

          <region>CA</region>

          <code>94089</code>

          <country>USA</country>
        </postal>

        <email>hannes@juniper.net</email>
      </address>
    </author>

    <author fullname="Bruno Decraene" initials="B" surname="Decraene">
      <organization>France Telecom - Orange</organization>

      <address>
        <postal>
          <street>38 rue du General Leclerc</street>

          <city>Issy Moulineaux cedex 9</city>

          <code>92794</code>

          <country>France</country>
        </postal>

        <email>bruno.decraene@orange.com</email>
      </address>
    </author>

    <date day="21" month="July" year="2014"/>

    <area>Routing</area>

    <workgroup>Network Working Group</workgroup>

    <keyword>2547</keyword>

    <keyword>l3vpn</keyword>

    <keyword>protection</keyword>

    <keyword>local repair</keyword>

    <keyword>fast reroute</keyword>

    <abstract>
      <t>This document specifies a fast-protection mechanism for protecting
      <xref pageno="false" target="RFC2547"/> based VPN service against egress
      node failure. This mechanism enables local repair to be performed
      immediately upon a egress node failure. In particular, the routers
      upstream to egress node could redirect VPN traffic to a protector (a new
      role) to repair in the order of tens of milliseconds, achieving fast
      protection that is comparable to MPLS fast reroute.</t>
    </abstract>
  </front>

  <middle>
    <section title="Introduction">
      <t>This document specifies a fast-protection mechanism for protecting
      RFC 2547 based VPN against egress PE failure. The procedures in this
      document are relevant only when a VPN site is multi-homed to two or more
      PEs. This is mainly designed based on MPLS context specific label
      switching<xref target="RFC5331"/>. This fast-protection refers to the
      ability to provide local repair upon a failure in the order of tens of
      milliseconds, which is comparable to MPLS fast-reroute <xref
      target="RFC4090"/>. This fast-protection is achieved by establishing
      local protection as close to a failure as possible. Compared with the
      existing global repair mechanisms that rely on control plane
      convergence, these procedures could provide faster and more
      deterministic restoration for VPN traffic. However, this is intended to
      complement the global repair mechanisms, rather than replacing them in
      any way.</t>
    </section>

    <section title="Specification of Requirements">
      <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>
    </section>

    <section title="Terminology">
      <t>Protected PE: A PE which request fast-protection for set of VPN-IP
      prefixes.</t>

      <t>Protected VPN-IP prefix: A multi-homed VPN-IP prefix that required
      protection in event of protected node goes down.</t>

      <t>Protector: A router which protect one or more Protected VPN-IP prefix
      when a Protected node goes down.</t>

      <t>BGP nexthop: A nexthop advertised in the BGP-Update for the VPN-IP
      prefix by a BGP speaker.</t>

      <t>VPN label: A label advertised by a BGP speaker for set of VPN-IP
      prefixes. This label could be per-VRF label or per-nexthop label or
      per-prefix label.</t>

      <t>Transport LSP: A MPLS LSP setup to BGP nexthop either by LDP or
      RSVP.</t>

      <t>Alternative egress PE: A PE originates VPN-IP prefix with same IP
      prefix of the protected VPN-IP prefix in a same VPN.</t>

      <t>Context MPLS table: A context-specific label space FIB. This table is
      populated with VPN labels advertised by the protected-PE for the
      protected VPN-IP prefix.</t>

      <t>Context label: A label from protector provides context for
      context-specific label forwarding.</t>

      <t>Context VRF: A IP FIB with alternate nexthop per context per
      site.</t>

      <t>PLR: Point of Local Repair.</t>
    </section>

    <section title="Reference topology">
      <t>This document refers to the following topologies to describe various
      roles, procedures and solution.</t>

      <figure align="center" anchor="Topology-1">
        <artwork align="center">


                  .......................
                  .                     .
   +-------+--CE1----PE1            PE4----CE5---+-------+
   | red   |      .     \           /   .        | red   |
   | site1 |      .      \         /    .        | site2 |
   +-------+--CE2-----+   P--P--PLR1  +----CE6---+-------+
                  .   |  /   |   | \  | .
                  .   PE2    RR  |  PE5 .
                  .   |  \   |   | /  | .
   +-------+--CE3-----+   P--P--PLR2  +----CE7--+-------+  
   | blue  |      .      /        \     .       |blue   |
   | site1 |      .     /          \    .       |site2  |  
   +-------+--CE4-----PE3           PE6----CE8--+-------+
                  .                     .
                  .                     .
                  .......................
           </artwork>
      </figure>

      <t>In <xref pageno="false" target="Topology-1"/> there are two VPNs red
      and blue with two multi-homed sites connecting to their PEs. Assume blue
      VPN site2 and red VPN site2 required egress protection in case of PE5
      goes down. Then PE5 is protected PE for red VPN site2 for and blue VPN
      site2. VPN-IP prefixes originated by PE5 associated with red site2 and
      blue site2 are protected VPN prefixes. The MPLS label associated with
      VPN-IP prefix is VPN Label. The PE4 is an alternative egress PE for red
      site2 and PE6 is an alternative egress PE for blue site2. The protector
      role could be delegated to any existing router in the network. For
      example PE4 could act as protector for red VPN site2 and PE6 could acts
      as protector for blue VPN site2. This protector model is co-located
      model. Alternatively, RR or any other router participates in VPN-IP
      control plane and not connected to VPN sites could also act as protector
      for both red and blue VPN site2. This model is centralized model.</t>

      <figure align="center" anchor="Topology-2">
        <artwork> 
                  .......................
                  .                     .
   +-------+--CE1----PE1            PE4----CE5---+-------+
   | red   |      .     \           /   .        | red   |
   | site1 |      .      \         /    .        | site2 |
   +-------+--CE2-----+   P--P--PLR1  +----CE6---+-------+
                  .   |  /   |   | \  | .
                  .   PE2    RR  |  PE5 .
                  .   |  \   |   | /  | .
   +-------+--CE3-----+   P--P--PLR2  +----CE7--+-------+  
   | red   |      .      /        \     .       |red    |
   | site3 |      .     /          \    .       |site4  |  
   +-------+--CE4-----PE3           PE6----CE8--+-------+
                  .                     .
                  .                     .
                  .......................
</artwork>
      </figure>

      <t>In <xref pageno="false" target="Topology-2"/> there is a VPN red with
      four sites and all sites are multi homed to their PEs. Assume site2 and
      site4 require egress protection in case PE5 goes down. Then PE5 is the
      protected PE for site2 and site4. PE4 and PE6 are alternate PEs for
      site2 and site4 respectively. Here also the protector role could be
      delegated to any existing router in the network. For example PE4 could
      act as a protector for site2 and PE6 could act as a protector for site4.
      This is called the 'co-located model'. Also PE4 or PE6 could act as
      protector for both sites. This is called the 'hybrid model'.</t>

      <t>The various protector models and deployment guidance are spelled-out
      in <xref pageno="false" target="Protector"/> and <xref pageno="false"
      target="Deployment"/>.</t>
    </section>

    <section anchor="theory" title="Theory of Operation">
      <t>Each (egress) PE attached to a given multi-homed site originates VPN-
      IP route(s) associated with the destination(s) within that site. Each
      such route should have its own Route Distinguisher, and its own
      next-hop, although all these routes have the same Route Target(s). Each
      (ingress) PE attached to other sites within the same VPN, import these
      route(s) into VRF creating more than one possible path to multi-homed
      sites. When an egress PE goes down, all VPN traffic destined to the
      multi homed sites attached to the downed egress PE gets rerouted to
      alternate egress PE(s) attached to same multi-homed site by ingress
      PE(s) after it detects the egress PE down. Until ingress PE(s) reroute
      the VPN traffic, the traffic that used to go through the failed PE get
      dropped in penultimate hop router. Even though connectivity of
      multi-homed site is not bound to an egress PE, the VPN traffic gets
      dropped in the P router as a result of the downed transport LSP that
      binds to that egress PE. This document specifies a mechanism that
      repairs VPN traffic at the point of failure (typically a P router which
      is penultimate hop of the transport LSP) and still keep P router unaware
      of the VPN information with the help of a protector. <xref
      target="Protector"/> explain the details. The penultimate hop router(s)
      of the transport LSP to egress PE(PLR) reroutes VPN traffic to protector
      through a bypass LSP in the event of egress PE failure. Protector
      forwards VPN traffic received from PLR in the bypass LSP to the
      alternative egress PE until the ingress PE reroute traffic to alternate
      egress PE.</t>

      <section anchor="Protector" title="Protector and Protection Models">
        <t>Protector, a new role, could be delegated to a router which
        participates in VPN-IP control plane for VPN-IP prefixes that requires
        egress node protection. In a network, protector could be the alternate
        egress PE of a egress protected multi homed site (precisely: the
        egress protected VPN-IP prefixes), or any other PE or stand-alone
        router for egress protection.</t>

        <t>This specification defines three types of protector:</t>

        <t><list style="symbols">
            <t>co-located</t>

            <t>centralized</t>

            <t>hybrid</t>
          </list>Its designation is dependent on the protector having direct
        links to the alternate site for a given VPN. A network MAY use either
        protection model or a combination depending on the requirements and
        actual network topology.</t>

        <section title="Co-located protector">
          <t>In this model, the protector role is delegated to the alternate
          egress PE for a protected VPN site. Protector is co-located with the
          alternate PE for the protected VPN site, and it has a direct
          connection to the multi-homed site that originates the protected
          VPN-IP prefix. In the event of an egress node failure, the protector
          receives traffic from the PLR, and forwards VPN traffic to the
          multi-homed site. In the <xref pageno="false" target="Topology-1"/>
          co-located protector could be PE4 red VPN site2 and PE6 could be the
          co-located protector for blue VPN site2.</t>
        </section>

        <section title="Centralized protector">
          <t>In this model, the protector serves as a centralized protector
          and does not have a direct connection to egress protected
          multi-homed sites. This model can be played by existing PEs or a
          dedicated protector. In the event of an egress PE failure, protector
          MUST forwards the traffic to an alternate egress PE with the VPN
          label advertised by the alternate egress PE for the VPN-IP prefix,
          which in turn forwards the traffic to the multi-homed site. In the
          <xref pageno="false" target="Topology-1"/>RR could act as protector
          for red's site2 and blue's site2 or PE6 could act as protector for
          red's site2 and PE4 acts as protector for VPN blue's site2. This is
          centralized protector model (A PE protecting VPN(s) and not
          connected to any protected VPN site).</t>
        </section>

        <section title="Hybrid protector">
          <t>In this model, the protector is co-located for some egress
          protected sites and centralized for other egress protected sites.
          These protected egress sites could be in the same VPN or in
          different VPN. In the <xref pageno="false"
          target="Topology-2"/>either PE4 or PE6 could act as hybrid
          protector. <xref pageno="false" target="Topology-1"/>PE6 could act
          as hybrid protector for VPNs red site2 and blue site2.</t>
        </section>
      </section>

      <section title="Context Identifier and VPN prefixes.">
        <t>Context-identifier is an IP address that is either globally unique
        or unique in the private address space of the routing domain. A
        context-identifier is shared between protected PE and protector(s) and
        It provides forwarding context for protected PE and protector. In the
        Protected PE each VPN-IP prefix is assigned to a context-identifier.
        The granularity of a context identifier is {Egress PE, VPN-IP prefix}
        tuple. However, a given context identifier MAY be assigned to one or
        multiple VPN-IP prefixes. A given context identifier MUST NOT be used
        by more than one protected PE and should never used for setting up BGP
        sessions or any control plane sessions.</t>

        <t>The egress PE that requires protection for a VPN-IP prefix MUST set
        context-identifier as the BGP nexthop for VPN-IPv4 and IPv4-Mapped
        context-identifier for VPN-IPv6. This context-identifier as nexthop
        indicates to the protector that a particular VPN-IP prefix need
        protection. For example in <xref target="Topology-1"/> PE5 (protected
        PE) advertises VPN-IP prefixes with context-identifier as BGP nexthop.
        The context identifier MUST also be advertised in the IGP and in LDP
        if LDP is used to establish transport LSP.</t>

        <t>Possible context identifier assignments are <list style="symbols">
            <t>Unique context-identifier for all VPN-IP prefixes, both
            VPN-IPv4 and VPN-IPv6. Here all the VRFs on a PE share same
            context-identifier.</t>

            <t>Unique context-identifier per address family. Here all the VRFs
            on the PE share the same context-identifier for given address
            family.</t>

            <t>Unique context-identifier per site for all VPN-IP prefixes,
            both VPN-IPv4 and VPN-IPv6. Here every VRFs has different
            context-identifier.</t>

            <t>Unique context-identifier per site per address family. Here
            every VRFs has different context-identifiers for a given address
            family.</t>

            <t>Unique context-identifier per CE address (nexthop). Here every
            CE in a VRF has a different context-identifier.</t>

            <t>Unique context identifier for each VPN-IP prefix. Here every
            VPN-IP has a different context-identifier.</t>
          </list></t>

        <t>The first one is coarsest granularity of a context identifier and
        the last one is finest granularity of a context identifier. While all
        of the above options are possible in principle, their practical usage
        is likely to vary, as not all of them may be of practical usage.</t>
      </section>

      <section anchor="MPLS-egress-frr" title="MPLS egress Fast reroute">
        <t>A Protector should be able to receive the traffic from PLR in the
        event of an egress PE failure with forwarding context that enables
        protector to repair VPN traffic.</t>

        <section anchor="RSVP-egress-frr" title="RSVP">
          <t>If RSVP LSP is used for transport then protector and primary MUST
          follow procedures specified in <xref target="rsvp-egress-frr">
          </xref>. The context-identifier will be used as destination address
          of the protected LSP and the protector will be backup egress node of
          the protected LSP. PLR MUST follow <xref target="rsvp-egress-frr"/>
          procedure if alias method is used.</t>
        </section>

        <section anchor="LDP-egress-frr" title="LDP">
          <t>If LDP is used for transport then LDP FEC MUST be the context
          identifier. The protector for the context identifier and context
          label could be learned through IGP which is beyond the scope of the
          document. The node protecting bypass path could be computed either
          by remote LFA or LFA for the context identifier to protector. This
          bypass LSP to protector with context label, learned through IGP,
          provide forwarding context to protector.</t>
        </section>
      </section>

      <section anchor="forwarding" title="Forwarding State on Protector PE">
        <t>A Protector MUST maintain multiple forwarding tables. Protector
        maintains the forwarding state in context-specific label space on per
        context-Identifier basis. It also maintains context specific IP
        forwarding table, context VRF, populated by extracting IP from VPN-IP
        prefix with nexthop to alternate egress PE for egress protected
        prefixes. In particular, the protector MUST learn VPN labels
        associated with VPN-IP prefixes by participating in VPN routing and
        MUST keep routes and labels associated with VPN(s) site(s) that
        required protection. For each VPN label with an associated
        context-identifier, the protector MUST map the context identifier to a
        context-specific label space <xref pageno="false" target="RFC5331"/>,
        and programs the VPN label in that label space into its forwarding
        plane. The VPN label in the context-specific label space identifies
        the IP forwarding table, that need to be looked up to send it
        alternate egress PE.</t>

        <t>The protector MAY maintain only VPN-IP prefix originated with-in
        the multi-homed site for given {egress PE, VPN} tuple. These VPN
        labels in context table and context VRF will not be used in forwarding
        after the ingress PE reroutes the traffic to the new best PE.
        Protector MUST delete VPN label and the VPN context table after
        ingress reroute the traffic. This SHOULD be achieved with a timer.
        This timer default value is 180 seconds, allowing to be able to
        sustain large reroute events.</t>

        <t>Note that if the protected PE does advertise a distinct label per
        VPN-IP prefix, as an optimization, the protector PE does not need to
        create an context VRF as the MPLS lookup on the VPN label is enough to
        identify the outgoing PE and label.</t>

        <section title="Alternate egress PE for protected prefix.">
          <t>Any route with BGP nexthop which has the following properties
          <list>
              <t>Exact matching route-target set</t>

              <t>Exact matching Prefix part (excluding the RD)</t>
            </list></t>

          <t>will be eligible as alternate egress PE for prefix.</t>
        </section>
      </section>
    </section>

    <section anchor="PE_failure" title=" Egress node Failure">
      <t>This section summarizes the procedure for egress protection as
      described in the above section for completeness. A Egress PE, Protector,
      PLR follows the methods described in <xref target="MPLS-egress-frr"/>.
      The protector programs forwarding state in such a way that packets
      received on the bypass LSP will be forwarded based on VPN label in the
      context table, and prefix lookup in context VPN table. The context table
      is identified by the UHP label of the bypass LSP, i.e. the context
      identifier.</t>

      <t>When the penultimate Hop router receives a VPN packet from the MPLS
      network, if the egress PE is down, the PLR tunnels the packet through
      the bypass LSP to the protector. The protector PE identifies the
      forwarding context of the egress PE based on the top label of the packet
      which is the UHP label of the bypass LSP. The protector further performs
      a second label lookup in the protected PE's context label space followed
      by layer-3 lookup in the VPN context table. These UHP label, context
      table label and layer-3 lookup results in forwarding the packet to the
      site or send it to alternate egress PE based on protector model.</t>

      <t>For example in <xref target="Topology-1"/> RR acts as Protector and
      PE5 requires protection for red, blue site2 VPN-IP prefixes. As red
      site2 and blue site2 VPN-IP prefixes are advertised with
      context-identifier, the protector sets up the forwarding table for
      VPN-IP prefixes from site2 with alternative egress PE as nexthop. When
      PLR detects PE5 failure it sends all the traffic that PLR used to
      forward directly to PE5 to protector through bypass LSP. In the
      protector the top label identifies the context specific table. The VPN
      label in the context table identifies the VPN layer-3 forwarding table
      which contains site2 VPN-IP prefixes with alternate PE as nexthop. A
      Layer-3 lookup gives mpls path to alternate egress PE and protector will
      forward the packet to alternate egress PE and reach to the site2.</t>
    </section>

    <section anchor="Deployment" title="Deployment Considerations">
      <section title="Discussion on deployment models.">
        <t>As the context-identifiers are advertised in the IGP, they
        introduce additional states in the network and the forwarding tables.
        As such, in general, it's desirable to keep their number limited. The
        granularity of context-identifier is also related to the protector
        model used. If a centralized or hybrid protector model is used, a
        unique context-identifier per egress PE is enough. If a co-located
        protector model is used, a context identifier per VPN or per CE may be
        needed.</t>

        <t>The centralized protector model, using a single context identifier
        per protected PE, limits the number of additional states in the
        network (IGP, forwarding tables) but may add extra latency during the
        protection time. It also minimizes the configuration effort as zero
        configuration is achievable. On the contrary the co-located mode,
        having a more granular context identifier, will minimize the latency
        during the protection time at the cost of adding more states in the
        network. It requires more configuration as the service provider will
        need to defines the PE pairs (protected, protector). The hybrid model
        is expected to offer the best trade-off as the number of IGP states in
        the network can be minimized by using a single context identifier per
        protected PE, while the additional latency can be limited by
        geographically distributing the protector PE in the network.</t>
      </section>

      <section title="Simple deployment model.">
        <t>We propose the following simple deployment model: <list
            style="symbols">
            <t>a single centralized Protector PE.</t>

            <t>a single context-identifier per protected PE, with all VPN
            routes advertised with this context-identifier as BGP
            next-hop.</t>
          </list></t>

        <t>It provides the following benefits: <list style="symbols">
            <t>minimize the number of IGP states in the network.</t>

            <t>minimize the configuration required: no per VPN configuration
            on the protector PE.</t>
          </list></t>

        <t>Regarding the IGP states, no additional states are required if the
        PEs uses secondary loopback address as BGP nexthop for VPN-IP address
        family. Otherwise, one additional IP address per PE need is needed.
        However, the number of IP address used as BGP next-hop for the
        customer traffic is not increased, hence if the routers allows the
        prioritization of the prefix during FIB update, there is no impact on
        the IGP convergence time.</t>

        <t>Regarding the configuration required on the network: <list
            style="symbols">
            <t>The protected PE is configured once with an additional IP
            address which serves as a context identifier. The BGP Next-Hop of
            the BGP routes are set to this context-identifier.</t>

            <t>The centralized protector PE does not required per VPN
            configuration. But it should allow set of context-identifiers to
            control VPN or PE it need to protects. This will be useful in
            multiple protectors in the network and set of PEs are protected by
            a given protector. The configured context-identifiers in protector
            protects subset of sites or PEs.</t>
          </list></t>

        <t>If one want to limit the protection to only a subset of VPN or a
        subset of PE (for lower VPN-SLA reasons, FIB capacities reasons on the
        protector, forwarding capacity reason during the protection time, for
        the hybrid model), one may not set context-identifier as a nexthop to
        the VPN-IP routes that required protection. VPN per protected PE
        configuration is required if user wants to limit egress protection for
        subset of sites. In this case protected be should allow user to not
        set the context-identifier as BGP nexthop for advertised VPN-IP
        prefixes.</t>
      </section>

      <section title="Deployment requirements.">
        <t>This solution does not mandate any protocol extension on any
        router. It does not mandate any additional feature on any routers
        except the new protector PE. In particular, it does not mandate
        implementation change on ingress nor egress PE, hence could works with
        legacy PE. In most topology, when LDP is used, the PLR will need to
        support the use of a LDP LSP as a targeted LFA. This is similar to
        R-LFA but the ability to configure a specific LSP to reach the
        protector PE may be specific.</t>
      </section>
    </section>

    <section anchor="Security" title="Security Considerations">
      <t>The security considerations discussed in RFC 5036, RFC 5331, RFC
      3209, and RFC 4090 apply to this document.</t>
    </section>

    <section anchor="ack" title="Acknowledgements">
      <t>This draft is based on the ideas originally developed by JL Le Roux,
      Bruno Decraene and Zubair Ahmad. This document leverages work done by
      Yakov Rekhter and several others on LSP tail-end protection. Thanks to
      Nischal Sheth, Nitin Bahadur, Yimin shen, Kaliraj Vairavakkalai and
      Maciek Konstantynowicz for their valuable contribution.</t>
    </section>
  </middle>

  <back>
    <references title="Normative References">
      &RFC5331;

      <?rfc include="reference.RFC.2547"?>

      <?rfc include="reference.RFC.4364"?>

      <?rfc include="reference.RFC.5036"?>

      <?rfc include="reference.RFC.2205"?>

      <?rfc include="reference.RFC.3209"?>

      <?rfc include="reference.RFC.4090"?>

      <?rfc include="reference.RFC.3471"?>

      <?rfc include="reference.RFC.3031"?>

      <reference anchor="LDP-UPSTREAM">
        <front>
          <title>MPLS Upstream Label Assignment for LDP</title>

          <author fullname="Rahul Aggarwal" initials="R" surname="Aggarwal">
            <organization>Juniper Networks</organization>
          </author>

          <author fullname="Jean-Louis Le Roux" initials="J. L. Le"
                  surname="Roux">
            <organization>France Telecom</organization>
          </author>

          <date year="2011"/>
        </front>

        <seriesInfo name="Internet-Draft" value="draft-ietf-mpls-ldp-upstream"/>

        <format target="http://tools.ietf.org/html/draft-ietf-mpls-ldp-upstream-10"
                type="TXT"/>
      </reference>

      <reference anchor="RSVP-NON-PHP-OOB">
        <front>
          <title>Non PHP Behavior and out-of-band mapping for RSVP-TE
          LSPs</title>

          <author fullname="Zafar Ali" initials="A" surname="Ali">
            <organization>Cisco Systems, Inc</organization>
          </author>

          <author fullname="George Swallow" initials="Z" surname="Swallow">
            <organization>Cisco Systems, Inc</organization>
          </author>

          <author fullname="Rahul Aggarwal" initials="R" surname="Aggarwal">
            <organization>Juniper Networks</organization>
          </author>

          <date year="2011"/>
        </front>

        <seriesInfo name="Internet-Draft"
                    value="draft-ietf-mpls-rsvp-te-no-php-oob-mapping"/>

        <format target="http://tools.ietf.org/html/draft-ietf-mpls-rsvp-te-no-php-oob-mapping-07"
                type="TXT"/>
      </reference>
    </references>

    <references title="Informative References">
      <?rfc include="reference.RFC.5920"?>

      <?rfc include="reference.RFC.5286"?>

      <?rfc include="reference.RFC.5714"?>

      <reference anchor="rsvp-egress-frr" target="rsvp egress frr">
        <front>
          <title>IP Fast Reroute Framework</title>

          <author fullname="Jeyananth Minto Jeganathan" initials="J"
                  surname="Jeganathan"/>

          <author fullname="Hannes Gredler" initials="H" surname="Gredler"/>

          <author fullname="Yimin Shen" initials="Y" surname="Shen"/>

          <date month="Oct " year="2012"/>
        </front>

        <seriesInfo name="Internet-Draft"
                    value="draft-minto-rsvp-lsp-egress-fast-protection-01"/>

        <format target="http://tools.ietf.org/html/draft-minto-rsvp-lsp-egress-fast-protection-01"
                type="TXT"/>
      </reference>
    </references>
  </back>
</rfc>

PAFTECH AB 2003-20262026-04-22 17:33:53