INTERNET-DRAFT                                               R. Fernando
Intended Status: Standards track                                  D. Rao
Expires: April 28, 2015                                            Cisco
                                                                 L. Fang
                                                            M. Napierala
                                                                   N. So
                                                           Vinci Systems
                                                               A. Farrel
                                                        Juniper Networks

                                                        October 25, 2014

      Virtual Topologies for Service Chaining in BGP/IP MPLS VPNs



   This document presents techniques built upon BGP/IP MPLS VPN control
   plane mechanisms to construct virtual topologies for service
   chaining. These virtual service topologies interconnect network zones
   and constrain the flow of traffic between these zones via a sequence
   of service nodes so that service functions can be applied to the

   This document also describes approaches enabled by both the routing
   control plane and by network orchestration to realize these virtual
   service topologies.

Status of this Memo

   This Internet-Draft is submitted to IETF in full conformance with the
   provisions of BCP 78 and BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at

Fernando, et. al.        Expires April 28, 2015                 [Page 1]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   The list of Internet-Draft Shadow Directories can be accessed at

Fernando, et. al.        Expires April 28, 2015                 [Page 2]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

Copyright and License Notice

   Copyright (c) 2014 IETF Trust and the persons identified as the
   document authors. All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   ( in effect on the date of
   publication of this document. Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document. Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
     1.1  Terminology . . . . . . . . . . . . . . . . . . . . . . . .  4
   2.  Intra-Zone Routing and Traffic Forwarding. . . . . . . . . . .  5
   3.  Inter-Zone Routing and Traffic Forwarding. . . . . . . . . . .  7
     3.1  Traffic Forwarding Operational Flow . . . . . . . . . . . .  8
   4.  Inter-Zone Model . . . . . . . . . . . . . . . . . . . . . . .  9
     4.1  Constructing the Virtual Service Topology . . . . . . . . .  9
     4.2 Per-VM Service Chains. . . . . . . . . . . . . . . . . . . . 12
   5.  Routing Considerations . . . . . . . . . . . . . . . . . . . . 12
     5.1  Multiple Service Topologies . . . . . . . . . . . . . . . . 12
     5.2  Multipath . . . . . . . . . . . . . . . . . . . . . . . . . 12
     5.3  Supporting Redundancy . . . . . . . . . . . . . . . . . . . 12
     5.4  Route Aggregation . . . . . . . . . . . . . . . . . . . . . 13
   6. Orchestration Driven Approach . . . . . . . . . . . . . . . . . 13
   7.  Security Considerations. . . . . . . . . . . . . . . . . . . . 13
   8.  Management Considerations. . . . . . . . . . . . . . . . . . . 13
   9.  IANA Considerations. . . . . . . . . . . . . . . . . . . . . . 13
   10.  Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . 14
   11.  References. . . . . . . . . . . . . . . . . . . . . . . . . . 14
     11.1  Normative References . . . . . . . . . . . . . . . . . . . 14
     11.2  Informative References . . . . . . . . . . . . . . . . . . 14
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 15

Fernando, et. al.        Expires April 28, 2015                 [Page 3]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

1.  Introduction

   Network topologies and routing design in enterprise, data center, and
   campus networks typically reflect the needs of the organization in
   terms of performance, scale, security, and availability. For scale
   and security reasons, these networks may be composed of multiple
   small domains or zones each serving one or more functions of the

   A network zone is a logical grouping of physical assets that supports
   certain applications. Hosts can communicate freely within a zone.
   That is, a datagram traveling between two hosts in the same zone is
   not routed through any servers that examine the datagram payload and
   apply services (such as security or load balancing) to the traffic.
   But a datagram traveling between hosts in different zones may be
   subject to additional services to meet the needs of scaling,
   performance, and security for the applications or the networks

   Networks have achieved division into zones and the imposition of
   services through a combination of physical topology constraints and
   routing. For example, one can force datagrams to go through a
   firewall (FW) by putting the FW in the physical data path from a
   source to the destination, or by causing the routed path form source
   to destination to go via a FW that would not normally be on the path.
   Similarly, the datagrams may need to go through a security gateway
   for security services, or a Load Balancer (LB) for load balancing

   In virtualized data centers, appliances, applications, and network
   functions, including IP VPN provider edge (PE) and customer edge (CE)
   functions are all commonly virtualized. That is, they exist as
   software instances residing in servers or appliances instead of
   individual (dedicated) physical devices.

   Migrating a network with all its functions and infrastructure
   elements to realization in a virtualized data center requires network
   overlay mechanisms that provide the ability to create virtual network
   topologies that mimic physical networks, and that provide the ability
   to constrain the flow of routing and traffic over these virtual
   network topologies.

   A data center uses a virtual topology in which the servers are in the
   "virtual" data path, rather than in the physical data path. For
   example, a traffic flow might previously have had the source PE-1 and
   destination at an Autonomous System Border Router (ASBR), ASBR-1, and
   the flow might have needed to be serviced by FW-1 and LB-1. In this
   virtualized data center, the functions of all four nodes could be

Fernando, et. al.        Expires April 28, 2015                 [Page 4]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   provided by virtual nodes that could be placed at arbitrary locations
   across the data center. Thus the "virtual service chain" vPE-1, FW-1,
   vLB-1, vASBR-1, that is the sequence of virtual service nodes that
   packet must traverse, could be realized by a logical path between
   arbitrary physical locations in the data center.

   A data center will likely support multiple tenants. A tenant is a
   customer who uses the virtualized data center services. Each tenant
   might require different connectedness (i.e., a different virtual
   topology) between their zones and applications, and might need the
   ability to apply different network policies such that the services
   for inter-zone traffic are applied in a specific order according to
   the organization objectives of the tenant. Furthermore, a data center
   might need multiple virtual topologies per tenant to handle different
   types of application traffic.

   Additionally, a data center operator may choose to provide services
   for multiple tenants on the same virtualized end device, for example,
   a server. Such multi-tenant devices must utilize techniques such as
   routing isolation to retain separation between tenants' traffic.

   To address all of these requirements, the mechanisms devised for use
   in a data center need to be flexible enough to accommodate the custom
   needs of the tenants and their applications, and at the same time
   must be robust enough to satisfy the scale, performance, and high
   availability needs that are demanded by the operator of the virtual
   network infrastructure that has a very large number of tenants each
   with different application types, large networks, multiple services,
   and high-volume traffic.

   Toward this end, this document introduces the concept of virtual
   service topologies and extends IP MPLS VPN control plane mechanisms
   to constrain routing and traffic flow over virtual service

   The creation of these topologies and the setting up of the forwarding
   tables to steer traffic over them may be carried out either by
   extensions to IP MPLS VPN procedures and functionality at the PEs, or
   via a "software defined networking" (SDN) approach. This document
   specifies the use of both approaches, but uses the IP MPLS VPN option
   to illustrate the various steps involved.

   This draft is a re-submission of draft-rfernando-l3vpn-service-
   chaining-05.txt, to conform to the bess working group nomenclature.

Fernando, et. al.        Expires April 28, 2015                 [Page 5]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

1.1  Terminology

   This document uses the following acronyms and terms.

   Terms             Meaning
   -----             --------------------------------------------------
   AS                Autonomous System
   ASBR              Autonomous System Border Router
   CE                Customer Edge
   FW                Firewall
   I2RS              Interface to the Routing System
   L3VPN             Layer 3 VPN
   LB                Load Balancer
   NLRI              Network Layer Reachability Information [RFC4271]
   P                 Provider backbone router
   proxy-arp         proxy-Address Resolution Protocol
   RR                Route Reflector
   RT                Route Target
   SDN               Software Defined Network
   vCE               virtual Customer Edge router
   vFW               virtual Firewall
   vLB               virtual Load Balancer
   VM                Virtual Machine
   vPC               virtual Private Cloud
   vPE               virtual Provider Edge router
   VPN               Virtual Private Network
   VRF               VPN Routing and Forwarding table [RFC4364]
   vRR               virtual Route Reflector

   This document also uses the following general terms:

     A BGP/IP MPLS VPN PE to which a service node in a virtual service
     topology is attached. The PE directs incoming traffic from other
     PEs or from attached hosts to the service node via an MPLS VPN
     label or IP lookup. The PE also forwards traffic from the service
     node to the next node in the chain. A Service-PE is a logical
     entity and a given PE may be attached to both a service node and
     an application host VM.

   Service node:
     A physical or virtual service appliance/application which inspects
     and/or redirects the flow of inter-zone traffic. Examples of
     service nodes include FWs, LBs, and deep packet inspectors. The
     service node acts as a CE in the VPN network.

Fernando, et. al.        Expires April 28, 2015                 [Page 6]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   Service chain: A sequence of service nodes that interconnect the
     zones containing the source and destination hosts or endpoints. The
     service chain is unidirectional and creates a one way traffic flow
     between source zone and destination zone.

   Virtual service topology:
     A virtual service topology consists of a sequence of service-PEs
     and their attached service nodes created in a specific order. A
     service topology is constructed via one or more routes that direct
     the traffic flow among the PEs that form the service chain.

     A BGP route attribute that identifies the specific service

     A tenant is a higher-level management construct. In the control/
     forwarding plane it is the collection of various virtual networks
     that get instantiated. A tenant may have more than one virtual
     network or VPN.

     A logical grouping of physical or virtual assets that supports
     certain applications or a subset thereof. VMs or hosts can
     communicate freely within a zone.

2.  Intra-Zone Routing and Traffic Forwarding

   This section provides a brief overview of how the BGP/IP MPLS VPN
   [RFC4364] control plane can be used in a DC network to used to divide
   the network into a number of zones. The subsequent sections in the
   document build on this base model to create inter-zone service
   topologies by interconnecting these zones and forcing inter-zone
   traffic to travel through a sequence of servers where the sequence of
   servers depends on the tuple <source zone, destination zone,

   The notion of a BGP/IP VPN when applied to the virtual data center
   works in the following manner.

   The VM that runs the applications in the server is treated as a CE
   attached to the VPN. A CE/VM belongs to a zone. The PE is the first
   hop router from the CE/VM and the PE-CE link is single hop from a
   layer-3 perspective. Any of the available physical, logical or
   tunneling technologies can be used to create this "direct" link
   between the CE/VM and its attached PE(s).

   If a PE attaches to one or more CEs of a certain zone, the PE must

Fernando, et. al.        Expires April 28, 2015                 [Page 7]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   have exactly one VRF for that zone, and the PE-CE links to those CEs
   must all be associated with that VRF. Intra-zone connectivity between
   CE/VMs that attach to different PEs is achieved by designating an RT
   per zone (zone-RT) that is both an import RT and an export RT of all
   PE VRFs that terminate the CE/VMs that belong to the zone. A VM may
   have multiple virtual interfaces that attach to different zones.

   It is further assumed that the CE/VMs are associated with network
   policies that are activated on an attached PE when a CE/VM is
   instantiated. These policies dictate how the network is set up for
   the CE/VM including the properties of the CE-PE link, the IP address
   of the CE/VM, the zones to which it belongs, QoS policies, etc. There
   are many ways to accomplish this step, but a description of such
   mechanisms is outside the scope of this document.

   When the CE/VM is activated, the attached PE starts to export the CEs
   IP address with the corresponding zone-RT. This allows unrestricted
   any-to-any communication between the newly active VM and the rest of
   the VMs in the zone.

   The classification of VMs into a zone is driven by the communication
   and security policy and is independent of the addressing scheme for
   the VMs. The VMs in a zone may be in the same or different IP subnets
   with user-defined mask-lengths. The PE advertises /32 routes to
   advertise reachability to locally attached VMs. If two VMs are in the
   same IP subnet, the PE may employ proxy-ARP to assist the VM to
   resolve ARP for other VMs in the IP subnet, and may use IP forwarding
   to carry traffic between the VMs. When a VM is attached to a remote
   PE, IP VPN forwarding is used to tunnel packets to the remote PE.

3.  Inter-Zone Routing and Traffic Forwarding

   A simple form of inter-zone traffic forwarding can be achieved using
   extranets or hub-and-spoke L3VPN configurations [RFC7024]. However,
   the ability to enforce constrained traffic flows through a set of
   services is non-existent in extranets and is limited in hub-and-spoke

   Note that the inter-zone services cannot always be assumed to reside
   and be in-lined on a PE. There is a need to virtualize the services
   themselves so that they can be implemented on commodity hardware and
   scaled out 'elastically' when traffic demands increase. This creates
   a situation where services for traffic between zones may be applied
   not only at the source-zone PE or the destination-zone PE. Mechanisms
   are required that make it easy to direct inter-zone traffic through
   the appropriate set of service nodes that might be remote or

Fernando, et. al.        Expires April 28, 2015                 [Page 8]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

3.1  Traffic Forwarding Operational Flow

   Traffic from a source endpoint (a VM/CE or PE) in a source zone
   reaches an ingress zone-PE and is associated with a VRF in that zone
   as described above. The zone-PE will forward the traffic and direct
   it toward the first service-node. If the service-node is attached to
   the zone-PE, the zone-PE will forward the packets out of one of its
   access interfaces. If the service-node is attached to a different
   service-PE, the zone-PE will encapsulate the packets and send them
   toward the service-PE. The zone-PE and service PE may be connected
   via an intermediate network of devices and the encapsulation causes
   the packets to be tunneled across this intermediate network.

   The service-PE will receive these encapsulated packets from the
   source zone-PE, decapsulate them, and forward them to its attached
   service-node. The traffic that comes back to the service-PE from the
   service-node must now be forwarded to the next service-node in the
   chain. As above, the next service-node may be locally attached or at
   a remote service-PE.

   At the last service-PE in the chain, the traffic that comes back from
   a service-node must be forwarded to the destination in the target
   zone. Just as with the service-nodes, the destination may be attached
   to the service-PE or reachable via another PE.

   As can be seen from this description, a given packet flow needs to be
   forwarded differently at each PE depending on whether it is arriving
   from a node attached to the PE or from a remote PE, and depending on
   whether the traffic is to be routed toward a node attached to the PE
   or attached to a remote PE. The next-hop for a flow changes depending
   on the relative position within the service chain.

   Figure 1 illustrates a virtual service topology, where hosts in Zone
   1 are interconnected with hosts in Zone 2 via two service nodes
   (Serv-A and Serv-B) attached to two service-PEs (S-PE-A and S-PE-B

   """"""""""""""""""""""                         """"""""""""""""""""""
   "          +-------+ " +--------+   +--------+ " +-------+          "
   " +-----+  | vPE-1 | " | S-PE-A |   | S-PE-B | " | vPE-2 |  +-----+ "
   " |VM/CE|--|       |---|        |---|        |---|       |--|VM/CE| "
   " +-----+  |(VRF-1)| " |(VRF-A) |   |(VRF-B) | " |(VRF-2)|  +-----+ "
   "          +-------+ " +--------+   +--------+ " +-------+          "
   "                    "      |            |     "                    "
   "     Zone 1         " +--------+   +--------+ "       Zone 2       "
   """""""""""""""""""""" | Serv-A |   | Serv-B | """"""""""""""""""""""
                          +--------+   +--------+

Fernando, et. al.        Expires April 28, 2015                 [Page 9]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   The different forwarding paths can be achieved at any PE as follows.

   o Each service node is associated with two VRFs at the service PE to
     which it is attached: an in-VRF for traffic toward the service
     node, and an out-VRF for traffic from the service node.

   o Traffic for the in-VRF arrives from the previous node in the
     service chain, and traffic for the out-VRF is destined toward the
     next node in the service chain, or toward the destination zone.

   o The in-VRF has one or more routes with a next-hop of a local access
     interface where the service node is attached. The out-VRF has
     routes with a next-hop of the next service node, which may be
     situated locally on the service-PE or at a remote PE.

   The installation of the forwarding entries to implement the flow
   described above may be achieved either via IP VPN mechanisms
   described in Sections 4 and 5, or using an SDN approach, as described
   in Section 6.

   It should be noted that the steps and constructs are logical, and may
   be implemented differently at each PE. Some options are specified in
   this document where pertinent.

4.  Inter-Zone Model

   The inter-zone model to realize the forwarding operational flow
   described in the previous section can be categorized into the
   following steps.

4.1  Constructing the Virtual Service Topology

   The virtual service topology described in the previous section is
   constructed via one or more service-topology routes that direct the
   traffic flow among the PEs and service nodes forming the service
   chain. There should be a route to reach each service node. The
   service topology, and hence the service routes, are constructed on a
   per-VPN basis. This service topology setup is typically independent
   of the routes for the actual destinations or flows that map to the
   service topology. There can be multiple service topologies for a
   given VPN.

4.1.1  Reachability to the Service Nodes

   Each service node is identified by an IP address that is scoped
   within the VPN. Specifically, there is an IP address for each
   interface on the node that is part of a service chain. The service
   node is also associated with an in-VRF and out-VRF table at the

Fernando, et. al.        Expires April 28, 2015                [Page 10]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   attached service PE.

   Reachability to the various service nodes in the service chain is
   achieved via regular BGP/IP VPN route advertisements.

   A service-PE will export a route to reach each service node interface
   attached to it. The service node interface may be a physical or
   logical interface. Each route will contain the Route-Target
   configured for the VPN, and a forwarding label that enables the
   service-PE to directly forward incoming traffic from the other PEs to
   the service node.

   The routes to reach the various service nodes are imported into and
   installed in each service out-VRF at a service-PE, as well as in the
   ingress zone VRF on an ingress zone PE.

   The in-VRF and out-VRF are conceptual entities and may be realized in
   different ways, as described further in this document. For instance,
   the forwarding label described above serves the purpose of the in-VRF
   in directing traffic from other PEs to an attached service node.
   Similarly, a per-interface policy-based-routing rule applied to an
   access interface serves to direct traffic coming in from attached
   service nodes.

4.1.2  Provisioning the Service Chain

   At each PE supporting a given VPN, the sequence of service nodes in a
   service chain can be specified as a VPN service route-policy.

   To create the service chain and give it a unique identity, each PE
   may be provisioned with a route-policy that contains the following
   tuple for every service chain that it belongs to:

      {Service-topology-name, Service-topology-RT, Service-node-

   where Service-node-Sequence is simply an ordered list of the service
   node IP addresses that are in the chain.

   Every service chain has a single unique service-topology-RT that is
   provisioned on all participating PEs.

   Each participating PE will also be provisioned with the tables,
   interfaces and other VPN configuration for the various zone and
   service VRFs needed on the PE based on attached nodes.

   At an egress zone PE, the corresponding zone-VRF will have a route-
   policy that attaches the appropriate Service-topology-RTs to routes

Fernando, et. al.        Expires April 28, 2015                [Page 11]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   exported from the VRF. The ingress zone and service VRFs will have
   the relevant Service-topology-RTs as import RTs.

   The service route-policy that lists the service chain tuple described
   above may be specified using various mechanisms. Using a YANG-defined
   data model is one suitable option.

4.1.3  Service-topology Next-Hop Resolution

   Routes representing hosts, VMs or other destinations associated with
   a zone are termed as zone prefixes. A zone prefix will have its
   regular zone-RTs attached when it is originated. This will be used by
   PEs that have VRFs for the same zone to import these prefixes to
   enable direct communication between end-points in the same zone.

   In addition to the intra-zone RTs, zone prefixes may also be tagged
   at the point of origination or an intermediate point, with the set of
   Service-topology-RTs corresponding to the service chains that have
   been set up towards this zone.

   Since they are tagged with the Service-topology-RT, zone prefixes can
   get imported into the VRFs of the PEs that form part of the service
   chain associated to that Service-topology-RT. These routes may be
   installed in the out-VRF at the service-PEs as well as in the ingress
   zone's VRF.

   However, the procedures described below introduce a change in the
   actions related to next-hop resolution and route installation in the
   service and ingress zone VRFs. These actions change the behavior of a
   PE compared to normal BGP VPN behavior, but does not mandate protocol
   changes to BGP. This modification to PE behavior allows the automatic
   and constrained flow of traffic via the service chain.

   The PE, based on the presence of a Service-topology-RT in the zone
   routes it receives, will perform the following actions:

   1. It will ignore the next-hop and VPN label that were advertised in
      the NLRI.

   2. Instead, it will select as next-hop the appropriate service node
      from the Service-node sequence provisioned for the Service-
      topology-RT. Specifically, in the out-VRF associated with each
      attached service node, it will select the next service node in the

   3. It will further resolve this service next-hop IP address locally
      in the associated VRF, instead of in the global routing table. The
      VRF routing table contains routes to reach the service node's IP

Fernando, et. al.        Expires April 28, 2015                [Page 12]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

      address, installed as per Sec.4.1.1. The PE will use the next-hop
      (and label, if remote) associated with this IP address to
      encapsulate traffic toward the next service node.

   4. If the importing service-PE is the last service-PE as per the
      service chain tuple, it will use the VPN next hop that was
      advertised with the zone prefix, for route resolution and
      installation. It will also use the VPN label that came with the

   In this way, the zone prefixes in the intermediate service-PE hops
   recurse over the nodes in the service chain forcing the traffic
   destined to them to flow through the virtual service topology.

   A significant benefit of this next-hop indirection is to avoid
   redundant advertisements of zone prefixes from the egress zones and
   service VRFs. Also, when the virtual service topology is changed (due
   to addition or removal of service nodes), there should be no change
   to the zone prefix's import/export RT configuration, and hence no re-
   advertisement of zone prefixes.

4.1.4  Zone Prefix Route Aggregation

   If the prefixes or addresses in a zone are aggregatable, instead of
   the individual zone host or prefix routes being imported and used at
   all hops along the chain, they may be aggregated at a specific hub PE
   and the aggregate zone prefix used along the service chain between
   zones. In such a case, the aggregate zone prefix will carry a
   service-topology-RT and get imported in the ingress zone and service

   In the simplest case, a default route may be used to carry a service-
   topology-RT and get imported into the various service VRFs, thereby
   setting up forwarding along the service topology. In this case,
   however, the actual zone prefixes will be attached with a separate
   service-topology-RT that is used only on the ingress zone-PE and the
   last service-PE to import the zone prefixes into the respective VRFs.

4.1.5  Fine-grained traffic steering

   In addition to a destination address or prefix, the steering of
   traffic into a service chain may also be based on attributes of the
   packet flow, such as source address or protocol and port types.
   [FLOWSPEC] is one option that can be used to direct specific VPN
   traffic into a specific service topology.

   In this case, it is a flow-spec route that is advertised with the
   appropriate Service-topology-RT attached, so that the importing PEs

Fernando, et. al.        Expires April 28, 2015                [Page 13]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   resolve it using the service chain tuple as described above and
   install the flow-spec route with the appropriate next-hop.

   Using the scheme described in section 4.1.4, the flow-spec route may
   be installed only in the ingress zone-PE VRF using a distinct

4.1.5  Multiple service topologies

   There should be one service topology RT per virtual service topology.
   There can be multiple virtual service topologies and hence service
   topology RTs in a given VPN.

   Virtual service topologies are constructed unidirectionally. Traffic
   in opposite directions between the same pair of zones will be
   supported by two different service topologies and hence two service
   topology routes. These two service topologies might or might not be
   symmetrical, i.e. they might or might not traverse the same sequence.

   As noted above, a service node route is advertised with a label that
   directs incoming traffic to the attached service node. Alternatively,
   an aggregate label may be used for the service route and an IP route
   lookup done in an in-VRF at the service-PE to send traffic to the
   service node.

   Note that a new service node could be inserted into the service chain
   seamlessly by just configuring the service policy appropriately.

4.2 Per-VM Service Chains

   While the service-topology-RT allows an efficient inheritance of the
   service chain for all VMs or prefixes in a zone, there may be a need
   to create a distinct service chain for an individual VM or prefix.
   This may be done by provisioning a separate service-topology RT and
   service node sequence. The VM route carries the service-topology RT,
   and the destination and service PEs are provisioned with this RT as
   described above.

5.  Routing Considerations

5.1  Multiple Service Topologies

   A service-PE can support multiple distinct service topologies for a

5.2  Service scaling

   One could use all tools available in BGP to constrain the propagation

Fernando, et. al.        Expires April 28, 2015                [Page 14]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   and resolution of state created by the service topology [RFC4684].

   Additional service nodes can be introduced to scale out a particular
   service. Each such service would be represented by a virtual IP
   address, and multiple service nodes associated with it. Multiple
   service-PEs may advertise a route to this address based on the
   presence of an attached service node instance, thereby creating
   multiple equal cost paths. This technique could be used to
   elastically scale out the service nodes with traffic demand.

5.3  Supporting Service Redundancy

   For stateful services an active-standby mechanism could be used at
   the service level. In this case, the inter-zone traffic should prefer
   the active service node over the standby service node.

   At a routing level, this is achieved by setting up two paths for the
   same service route: one path goes through the active service node and
   the other through the standby service node. The active service path
   can then be made to win over the standby service path by
   appropriately setting the BGP path attributes of the service topology
   route such that the active path succeeds in path selection. This
   forces all inter-zone traffic through the active service node.

6. Orchestration Driven Approach

   In an orchestration driven approach, there is no need for the zone or
   service PEs to determine the appropriate next-hops based on the
   specified service node sequence. All the necessary policy
   computations are carried out, and the forwarding tables for the
   various VRFs at the PEs determined, by a central orchestrator or

   The orchestrator communicates with the various PEs (typically virtual
   PEs on the end-servers) to populate the forwarding tables.

   The protocol used to communicate between the controller/orchestration
   and the PE/vPE must be a standard, programmatic interface. There are
   several possible options to this programmatic interface, some being
   under discussion in the IETF's Interface to Routing Systems (I2RS)
   initiative, [I-D.ietf-i2rs-architecture], [I-D.ietf-i2rs-problem-
   statement]. One specific option is defined in [IPSE].

7.  Security Considerations

   To be added.

8.  Management Considerations

Fernando, et. al.        Expires April 28, 2015                [Page 15]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

   To be added.

9.  IANA Considerations

   This proposal does not have any IANA implications.

Fernando, et. al.        Expires April 28, 2015                [Page 16]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

10.  Acknowledgements

   The authors would like to thank the following individuals for their
   review and feedback on the proposal: Eric Rosen, Jim Guichard, Paul
   Quinn, Peter Bosch, David Ward, Ashok Ganesan and Thomas Morin. The
   option of configuring an ordered sequence of service nodes via policy
   is derived from a suggestion from Eric Rosen.

11.  References

11.1  Normative References

   [RFC4364]  Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private
              Networks (VPNs)", RFC 4364, February 2006.

11.2  Informative References

   [RFC4271]  Rekhter, Y., Li, T., and S. Hares, "A Border Gateway
              Protocol 4 (BGP-4)", RFC 4271, January 2006.

   [RFC4684]  Marques, P., Bonica, R., Fang, L., Martini, L., Raszuk,
              R., Patel, K., and J. Guichard, "Constrained Route
              Distribution for Border Gateway Protocol/MultiProtocol
              Label Switching (BGP/MPLS) Internet Protocol (IP) Virtual
              Private Networks (VPNs)", RFC 4684, November 2006.

   [RFC7024]  Jeng, H., Uttaro, J., Jalil, L., Decraene, B., Rekhter,
              Y., and R. Aggarwal, "Virtual Hub-and-Spoke in BGP/MPLS
              VPNs", RFC 7024, October 2013.

              L. Fang, et al.,"BGP/MPLS IP VPN Virtual CE",
              draft-fang-l3vpn-virtual-ce, work in progress.

              L. Fang, et al., "BGP/MPLS IP VPN Virtual PE",
              draft-fang-l3vpn-virtual-pe, work in progress.

              Atlas, A., Halpern, J., Hares, S., Ward, D., and T Nadeau,
              "An Architecture for the Interface to the Routing System",
              draft-ietf-i2rs-architecture, work in progress.

              Atlas, A., Nadeau, T., and D. Ward, "Interface to the
              Routing System Problem Statement",
              draft-ietf-i2rs-problem-statement, work in progress.

Fernando, et. al.        Expires April 28, 2015                [Page 17]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

          Fernando, R., Boutros, S., Rao, D., "Interface to a
          Packet Switching Element",
          draft-rfernando-ipse-00, work in progress.

Fernando, et. al.        Expires April 28, 2015                [Page 18]

INTERNET DRAFT          Virtual Service Topology        October 25, 2014

Authors' Addresses

   Dhananjaya Rao
   170 W Tasman Dr
   San Jose, CA

   Rex Fernando
   170 W Tasman Dr
   San Jose, CA

   Luyuan Fang
   5600 148th Ave NE
   Redmond, WA 98052

   Maria Napierala
   200 Laurel Avenue
   Middletown, NJ 07748

   Ning So
   Vinci Systems, Inc.

   Adrian Farrel
   Juniper Networks

Fernando, et. al.        Expires April 28, 2015                [Page 19]