idnits 2.17.1 draft-ietf-l3vpn-end-system-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- == There are 11 instances of lines with private range IPv4 addresses in the document. If these are generic example addresses, they should be changed to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x, 198.51.100.x or 203.0.113.x. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (October 2, 2014) is 3494 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Downref: Normative reference to an Unknown state RFC: RFC 1027 ** Obsolete normative reference: RFC 5512 (Obsoleted by RFC 9012) -- Obsolete informational reference (is this intentional?): RFC 5575 (Obsoleted by RFC 8955) == Outdated reference: A later version (-11) exists of draft-ietf-mpls-in-udp-05 == Outdated reference: A later version (-11) exists of draft-ietf-l2vpn-evpn-08 Summary: 2 errors (**), 0 flaws (~~), 4 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group P. Marques 3 Internet-Draft Juniper Networks 4 Intended status: Standards Track L. Fang 5 Expires: April 5, 2015 Microsoft 6 N. Sheth 7 Juniper Networks 8 M. Napierala 9 AT&T Labs 10 N. Bitar 11 Verizon 12 October 2, 2014 14 BGP-signaled end-system IP/VPNs. 15 draft-ietf-l3vpn-end-system-04 17 Abstract 19 This document describes a solution in which the control plane 20 protocol specified in BGP/MPLS IP VPNs is used to provide a Virtual 21 Network service to end-systems. These end-systems may be used to 22 provide network services or may directly host end-to-end 23 applications. 25 Status of This Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at http://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on April 5, 2015. 42 Copyright Notice 44 Copyright (c) 2014 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 Table of Contents 59 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 60 1.1. Terminology . . . . . . . . . . . . . . . . . . . . . . . 3 61 2. Requirements . . . . . . . . . . . . . . . . . . . . . . . . 3 62 3. Applicability of BGP IP VPNs . . . . . . . . . . . . . . . . 4 63 4. Virtual network end-points . . . . . . . . . . . . . . . . . 7 64 5. VPN Forwarder . . . . . . . . . . . . . . . . . . . . . . . . 9 65 6. XMPP signaling protocol . . . . . . . . . . . . . . . . . . . 11 66 7. End-System Route Server behavior . . . . . . . . . . . . . . 17 67 8. Operational Model . . . . . . . . . . . . . . . . . . . . . . 18 68 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 69 10. Security Considerations . . . . . . . . . . . . . . . . . . . 21 70 11. XML schema . . . . . . . . . . . . . . . . . . . . . . . . . 22 71 12. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 23 72 13. References . . . . . . . . . . . . . . . . . . . . . . . . . 23 73 13.1. Normative References . . . . . . . . . . . . . . . . . . 24 74 13.2. Informational References . . . . . . . . . . . . . . . . 24 75 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 25 77 1. Introduction 79 This document describes the requirements for a network virtualization 80 solution that provides an IP service to end-system virtual 81 interfaces. It then discusses how the BGP IP VPNs [RFC4364] control 82 plane can be used to provide a solution that meets these 83 requirements. Subsequent sections provide a detailed discussion of 84 the control and forwarding plane components. 86 In BGP IP VPNs, Customer Edge (CE) interfaces connect to a Provider 87 Edge (PE) device which provides both the control plane and VPN 88 encapsulation functions required to implement a Virtual Network 89 service. This document decouples the control plane and forwarding 90 functionality of the PE device in order to enable the forwarding 91 functionality to be implemented in multiple devices. For instance, 92 the forwarding function can be implemented directly on the operating 93 system of application servers or network appliances. 95 1.1. Terminology 97 This document makes use of the following terms: 99 End-System: A compute node which primary function is to run 100 applications. It is assumed that end-systems support multiple 101 application instances (e.g. virtual-machines), each with its 102 independent network configuration. 104 End-System Route Server: A software application that implements the 105 control plane functionality of a BGP IP VPN PE device and a XMPP 106 server that interacts with VPN Forwarders. 108 Virtual Interface: An interface in an end-system that is used by a 109 virtual machine or by applications. It performs the role of a CE 110 interface in a BGP IP VPN network. 112 VPN Forwarder: The forwarding component of a BGP IP VPN PE device. 113 This functionality may be co-located with the virtual interface or 114 implemented by an external device. 116 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 117 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 118 document are to be interpreted as described in RFC 2119 [RFC2119]. 120 2. Requirements 122 Network Virtualization is used in both service provider as well as 123 enterprise networks to support multi-tenancy and network-based access 124 control. It may also be used to facilitate application instance 125 mobility. 127 Multi-tenancy allows a physical network to provide services to 128 multiple "customers" or "tenants", whether these are external 129 entities in the case of a Service Provider providing managed VPN 130 services or internal departments sharing an IT facility. Multi- 131 tenancy requires isolation of traffic and routing information between 132 tenants. 134 Within a tenant, it is often required to create multiple distinct 135 virtual networks, in order to be able to provide network-based access 136 control. In this service model, each virtual network behaves as a 137 "Closed User Group" (CUG) of virtual interfaces that are allowed to 138 exchange traffic freely, while traffic between virtual networks is 139 subject to access controls. This scenario can be found in both 140 enterprise campus networks, branch offices and data-centers. 142 It is often the case when network access control is used, that the 143 traffic patterns are such that there is significantly more traffic 144 crossing a CUG boundary than staying within such boundary. As an 145 example, in campus networks it is common to segregate users into CUGs 146 based on some classification such as the user's department. Campus 147 networks often see traffic patterns in which almost all the traffic 148 flows northbound to the data-center or internet boundaries. Similar 149 traffic patterns can be found in multi-tier applications in IT data- 150 centers. 152 Virtual interfaces are often configured to expect the concept of IP 153 subnet to match its closed user group. A network virtualization 154 solution should be able to provide this concept of IP subnet 155 regardless of whether the underlying implementation uses a multi- 156 access network or not. 158 Virtual interfaces should be able to directly access multiple closed 159 user groups without needing to traverse a gateway. Network access 160 policy should allow this access whether the source and destination 161 CUGs for a particular traffic flow belong to the same tenant or 162 different tenants. It is often the case that infrastructure services 163 are provided to multiple tenants. One such example is voice-over-IP 164 gateway services for branch offices. 166 Independently, but often associated with the previous two functions, 167 IP mobility is another network function that can be implemented using 168 network virtualization. By abstracting the externally visible 169 network address from the underlying infrastructure address, mobility 170 can be implemented without having to recur to home agents or large L2 171 broadcast domains. 173 IP Mobility requires the ability to "move" a virtual interface 174 without disrupting its TCP (or UDP) transport sessions. This 175 requires a mechanism that can efficiently communicate the mappings 176 between logical and physical addressing. 178 IP Mobility can be a result of devices physically moving (e.g., a 179 WiFi enabled laptop) or workload being diverted between physical 180 systems such as network appliances or application servers. 182 3. Applicability of BGP IP VPNs 184 BGP IP VPNs [RFC4364] is the industry de-facto standard for providing 185 "closed user group" functionality in WAN environments. It is used by 186 service providers in environments where several millions of routes 187 are present. It supports both isolated VPNs as well as overlapping 188 VPNs (often referred to as "extranets"). 190 The BGP IP VPN control plane has been designed to be able to 191 distribute the mapping between virtual address and location (next- 192 hop) to the subset of network nodes for which this information is 193 relevant, whenever that mapping changes. This provides an efficient 194 mechanism to address IP mobility requirements as compared to methods 195 that depend on a (cached) mapping request from the end-systems. 197 In its traditional usage in Service Provider networks, BGP IP VPN 198 functionality is implemented in a Provider Edge (PE) device that 199 combines both BGP signaling as well as VRF-based forwarding 200 functions. In practice, most PE devices in current use are multi- 201 component systems with the signaling and forwarding functionality 202 actually implemented in different processors attached to an internal 203 network. 205 This document assumes a similar separation of functionality in which 206 software appliances, the End-System Route Servers, implement the 207 control plane functionality of a PE device and a VPN Forwarder 208 implements the forwarding function usually found in a PE device 209 "line-card". The VPN Forwarder functionality may be co-located with 210 the end-system (e.g., implemented in the hypervisor switch or host OS 211 network drivers) or it may be external. For instance, residing in a 212 data-center switch or specialized appliance. 214 Operationally, BGP IP VPN technology has several important 215 characteristics: 217 It has a high-level of aggregation between customer interfaces and 218 managed entities (Provider Edge devices). 220 It defines VPNs as policies, allowing an interface to directly 221 exchange traffic with multiple VPNs and allowing for the topology 222 of the virtual network to be modified by modifying the policy 223 configuration. 225 It scales horizontally in terms of event propagation. By 226 increasing the number of signaling devices implementing the PE 227 control plane, it is possible to decrease the load on each 228 signaling device when it comes to propagating events that 229 originate in a specific location and must be propagated across the 230 network. 232 The last point is particularly relevant to the convergence 233 characteristics required for large scale deployments. BGP's 234 hierarchical route distribution capabilities allow a deployment to 235 divide the workload by increasing the number of End-System Route 236 Servers. 238 As an example consider a topology in which 100 End-System Route 239 Servers are deployed in a network each serving a subset of the VPN 240 forwarding elements. The Route Servers inter-connect to two top- 241 level BGP Route Reflectors [RFC4456]. 243 If an event (i.e. a VPN route change) needs to be propagated from a 244 specific end-system to 10,000 clients randomly distributed across the 245 network, each of the End-System Route Servers must generate 100 246 updates to its respective downstream clients. 248 By modifying this topology such that another 100 End-System Route 249 Servers are added, then each Route Server is now responsible to 250 generate 50 client updates. This example illustrates the linear 251 scaling properties of BGP: doubling the number of Route Servers (i.e. 252 the processing capacity) reduces in half the number of updates 253 generated by each (i.e. load at each processing node). 255 The same horizontal scaling techniques can be applied to the Route 256 Reflector layer in the example above by subsetting the VPN Route 257 space according to some pre-defined criteria (for instance VPN route 258 target) and using a pair of Route Reflectors per subset. 260 In the previous example we assumed a dense membership in which all 261 Route Servers have local clients that are interested in a particular 262 event. BGP also optimizes the route distribution for sparse events. 263 The Route Target Constraint [RFC4684] extension, builds an optimal 264 distribution tree for message propagation based on VPN membership. 265 It ensures that only the PEs with local receivers for a particular 266 event do receive it also decreasing the total load on the upstream 267 BGP speaker. 269 In the WAN environment, BGP IP VPN control plane scaling is focused 270 not primarily on route convergence times but on memory footprint of 271 embedded devices. While memory footprint does not have a similar 272 linear scaling behavior, memory technology available to software 273 appliances is often at 10x the scale of what is commonly found in WAN 274 environments. 276 The functionality present in the BGP IP VPN control plane addresses 277 the requirements specified in the previous section. Specifically, it 278 supports multiple potentially overlapping "groups", regular or "hub 279 and spoke" topologies and the scaling characteristics necessary. 281 The BGP IP VPN control plane supports not only the definition of 282 "closed user-groups" (VPNs in its terminology) but also the 283 propagation of inter-VPN traffic policies [RFC5575]. 285 Note that the signaling protocol itself is rather agnostic of the 286 encapsulation used on the wire as long as this encapsulation has the 287 ability to carry a 20 bit label. 289 Several network environments use a network infrastructure that is 290 only capable of providing an IP unicast service. In order to support 291 them, implementations of this document should support the MPLS in GRE 292 [RFC4023] encapsulation. Other encapsulations are possible, 293 including UDP based encapsulations [I-D.ietf-mpls-in-udp]. 295 4. Virtual network end-points 297 This document assumes that end-systems support one or more virtual 298 network interfaces in addition to a physical interface that is 299 associated with the underlying network infrastructure. Virtual 300 network interfaces can be associated with a restricted list of 301 applications via OS-dependent mechanisms, a Virtual Machine (VM), or 302 they can be used to provide network connectivity to all user 303 applications in the same way that a "VPN tunnel" interface is used to 304 provide access between an end-system (e.g., a laptop) and a remote 305 corporate network. 307 From an IP address assignment point of view, a virtual network 308 interface is addressed out of the virtual IP topology and associated 309 with a "closed user group" or VPN, while the physical interface of 310 the machine is addressed in the network infrastructure topology. 312 A virtual network interface is connected to a VPN Forwarder. This 313 VPN Forwarder MAY be co-located in the end-system or external. 315 Both static and dynamic IP address allocation can be supported. The 316 later assumes that the VPN Forwarder implements a DHCP relay or DHCP 317 proxy functionality. 319 Traffic that ingresses or egresses through a virtual network 320 interface is routed at the VPN Forwarder which acts as the first-hop 321 router (in the virtual topology). The IP configuration on the client 322 side of this virtual network interface (e.g., in the guest OS) can 323 follow one of two models: 325 point-to-point interface model. 327 multipoint interface model. 329 In a point-to-point interface model, the VPN client routing table 330 (e.g., on the guest OS) contains the following routing entries: a 331 host route to the local IP address, a host route to the first-hop 332 router via the virtual interface and a default route to the first-hop 333 router. This is the model typically used in "VPN tunnel" 334 configurations or other access technologies such as cable deployments 335 or DSL. When this model is used, the first-hop router IP address is 336 a link-local address that is the same on all first-hop routers across 337 a specific deployment. This first-hop IP address should not change 338 when a virtual interface moves between different machines. 340 In a multi-point interface model, the VPN client routing table (e.g., 341 on the guest OS) contains the following routing entries: a host route 342 to the local IP address, a subnet route to the local interface and 343 optionally a default route to a specific router address within that 344 subnet. In this model, the VPN client IP stack will issue address 345 resolution requests for any IP addresses it considers to be directly 346 attached to the subnet. The VPN Forwarder shall answer all address 347 resolution requests via Proxy ARP [RFC1027].The same technique is 348 applicable when Neighbor Discovery is used to resolve IPv6 addresses. 349 Address resolution request should be answered using a virtual MAC 350 address which SHOULD be the same across all VPN Forwarders in a 351 specific deployment. This virtual MAC address SHALL default to the 352 VRRP [RFC5798] virtual router MAC address for Virtual Router 353 Identifier (VRID) 1. 355 When the virtual topology first-hop router resides on the same 356 physical machine, the host OS is responsible to map the virtual 357 interface with a VPN specific routing table (without taking L2 358 addresses into consideration). In this case the mac-addresses known 359 to the guest OS are not used on the wire. 361 When the virtual topology first-hop router resides in an external 362 system (e.g., the first hop-switch) the virtual interface shall be 363 identified by the combination of the mac-address assigned to physical 364 interface of the end-system and a 802.1Q VLAN tag. The first-hop 365 switch should use a virtual router MAC address to answer any address 366 resolution queries. 368 Whenever an external VPN Forwarder is used and resiliency is desired, 369 the external VPN Forwarder should be redundant. It is desirable to 370 use VRRP as a mechanism to control the flow of traffic between the 371 end-system and the external VPN Forwarder. VRRP already defines the 372 necessary procedures to elect a single forwarder for a LAN. 374 This specification uses the VRRP virtual router MAC address as the 375 default L2 address for the VPN Forwarder as a client virtual 376 interface may move between locations where redundancy may not be 377 present. 379 While the VRRP Virtual Router MAC will be used to answer any address 380 resolution request made by the virtual interface client (e.g., the 381 guest VM) this does not imply that a single default router is elected 382 per virtual IP subnet. The ingress VPN Forwarder will perform an IP 383 forwarding decision based on the destination IP address of the 384 (payload) traffic. 386 VRRP router election is only relevant in selecting the VPN Forwarder 387 associated with a specific machine, when external forwarders are in 388 use. 390 5. VPN Forwarder 392 In this solution, the Host OS/Hypervisor in the end-system must 393 participate in the virtual network service. Given an end-system with 394 multiple virtual interfaces, these virtual interfaces must be mapped 395 onto the network by the guest OS such that applications on one 396 virtual interface cannot send traffic to networks they are not 397 authorized to communicate with or using source addresses not assigned 398 to the virtual interface. 400 When VPN forwarder functionality is implemented by the Host OS/ 401 Hypervisor, intermediate systems in the network do not require any 402 knowledge of the virtual network topology. This can simplify the 403 design and operation of the physical network. 405 When it is not possible or desirable to add the VPN forwarding 406 functionality to the end-system, it may be implemented by an external 407 system, typically located as close as possible to the end-system 408 itself. 410 Both models, co-located and external VPN Forwarder can co-exist in a 411 deployment. 413 In order to implement the BGP IP VPN Forwarder functionality a device 414 MUST implement the following functionality: 416 Support for multiple "Virtual Routing and Forwarding" (VRF) 417 tables; 419 VRF route entries map prefixes in the virtual network topology 420 to a next-hop containing a infrastructure IP address and a 421 20-bit label allocated by the destination Forwarder. The VRF 422 table lookup follows the standard IP lookup (best-match) 423 algorithm. 425 Associate an end-system virtual interface with a specific VRF 426 table; 427 When the the Forwarder is co-located with the end-system, this 428 association is implemented by an internal mechanism. When the 429 Forwarder is external the association is performed using the 430 mac-address of the end-system and a IEEE 802.1Q tag that 431 identifies the virtual interface within the end-system. 433 Encapsulate outgoing traffic (end-system to network) according to 434 the result of the VRF lookup; 436 Associate incoming packets (network to end-system) to a VRF 437 according to the 20-bit label contained in the packet; 439 The VPN Forwarder MAY support the ability to associate multiple 440 virtual interfaces with the same VRF. When that is the case, locally 441 originated routes, that is IP routes to the local virtual interfaces 442 SHALL NOT be used to forward outbound traffic (from the virtual 443 interfaces to the outside) unless a route advertisement has been 444 received that matches that specific IP prefix and next-hop 445 information. 447 As an example, if a given VRF contains two virtual interfaces, 448 "veth0" and "veth1", with the addresses 10.0.1.1/32 and 10.0.1.2/32 449 respectively, the initial forwarding state must be initialized such 450 that traffic from either of these interfaces does not match the 451 other's routing table entry. It may for instance match a default 452 route advertised by a remote system. Traffic received from other VPN 453 Forwarders, however, must be delivered to the correct local 454 interface. If at a subsequent stage a route is received from the 455 Route Server such that 10.0.1.2/32 has a next-hop with the IP address 456 of the local host and the correct label, the system may subsequently 457 install a local routing table entry that delivers traffic directly to 458 the "veth1" interface. This means that forwarding table entries 459 apply to downstream only by default. This capability can be used to 460 implement a hub-and-spoke topology, if required. 462 The 20-bit label which is associated with a virtual-interface is of 463 local significance only and SHOULD be allocated by the VPN Forwarder. 465 When an external VPN Forwarder is used the end-system MUST associate 466 each virtual interface with a VLAN [IEEE.802-1Q] that is unique on 467 the end-system. The switching infrastructure MUST be configured such 468 that multi-destination frames sourced from an end-system are only 469 delivered to VPN Forwarders used by this end-system and not to other 470 end-systems. 472 6. XMPP signaling protocol 474 End-System Route Servers must be aware of VPN membership on each 475 Forwarder as well as what IP addresses are currently associated with 476 each virtual interface. 478 VPN Forwarders must receive VPN route information from which to 479 populate their forwarding tables. External VPN Forwarders also need 480 to receive the virtual interface and IP address events from the end- 481 system for which they are VPN forwarders. In this case the end- 482 system assigns an 802.1Q VLAN tag to each virtual interface and 483 communicates that information to the Forwarder. 485 In order to exchange this information this specification uses the 486 XMPP [RFC6120] protocol along with the Publish-Subscribe [pubsub] 487 extension. 489 VPN forwarders (both co-located and external) establish XMPP sessions 490 with End-System Route Servers, acting as XMPP clients. When an 491 external VPN Forwarder is used, end-systems establish XMPP sessions 492 with VPN Forwarders. External VPN Forwarders act as XMPP servers for 493 end-systems which are associated with them. 495 A VPN Forwarder MAY connect to multiple End-System Route Servers for 496 reliability. In this case it SHOULD publish its information to each 497 of the Route Servers. It MAY choose to subscribe to VPN routing 498 information once only from one of the available gateways. 500 The information advertised by an XMPP client SHOULD be deleted after 501 a configurable timeout, when the session closes. This timeout should 502 default to 60 seconds. 504 +---------+ +--------+ 505 | RS | ----------- | BGP | 506 +---------+ +--------+ 507 // \ / 508 XMPP \ / 509 // \ / 510 +------------+ \ / 511 | end-system | \ / 512 +------------+ \/ 513 \\ /\ 514 XMPP / \ 515 \\ / \ 516 +---------+ / \ +--------+ 517 | RS | ----------- | BGP | 518 +---------+ +--------+ 520 The figure above represents a typical configuration in which an end- 521 system with a co-located VPN Forwarder is directly connected to two 522 End-System Routes Servers, which are in turn connected to multiple 523 BGP speakers which may be other L3VPN PEs or BGP route reflectors. 525 In deployment the number of End-System Route Servers used will depend 526 on the desired Route Server to VPN Forwarder ratio which affects the 527 convergence time of event propagation. 529 The XMPP "jid" used by the client shall be a string that uniquely 530 identifies it in its administrative domain. This specification 531 recommends the use of the hostname (when unique) or an IP address in 532 its string representation. 534 Each VPN shall be identified by a 128 octet ASCII character string. 536 When external Forwarders are used, its control software operates as a 537 XMPP server processing requests from end-systems and as a client of 538 one or more End-System Route Servers. The control software relays to 539 the End-System Route Server(s) VPN membership messages it receives 540 from the end-system. VPN routing information received from the Route 541 Server(s) SHOULD NOT be propagated to the end-system. 543 When a virtual interface is created on a end-system, the host 544 operating-system software shall generate an XMPP Subscribe message to 545 its server (the End-System Route Server or external VPN Forwarder). 547 Subscription request from co-located VPN Forwarder to Route Server: 549 553 554 555 556 1 557 558 559 561 The request above, instructs the End-System Route Server to start 562 populating the client's VRF table with any routing information that 563 is available for this VPN. The XMPP node 'vpn-customer-name' is 564 assumed to be implicitly created by the End-System Route Server. 565 Creation of a virtual interface may precede any IP address becoming 566 active on the interface, as it is the case with VM instantiation. 568 The optional "instance-id" element allows the VPN Forwarder to 569 specify a unique 16 bit index that can be used by the Route Server to 570 automatically assign a Route Distinguisher (RD) to any route 571 subsequently advertised by the VPN Forwarder. In a scenario where 572 the VPN Forwarder is advertising reachability information to multiple 573 Route Servers it is desirable for reachability information to have an 574 RD composed of the VPN Forwarder identifier (e.g. IPv4 address) and 575 the "instance-id". 577 Subscription request from end-system to external VPN Forwarder: 579 583 584 585 586 587 vlan-id 588 589 590 591 593 When an external VPN Forwarder is used, the end-system should include 594 the VLAN identifier it assigned to the virtual interface as a 595 subscription option. 597 When a IP address is added to a virtual interface, the end-system 598 will generate an XMPP Publish request. 600 Publish request from VPN Forwarder to End-System Route Server: 602 604 to='network-control@domain.org' 605 id='request1'> 606 607 608 609 610 611 1 612
'vpn-ip-address/32'
613 614 615 616 1 617
'infrastructure-ip-address'
618 619 620 gre 621 udp 622 623 624 625 1 626 627 628 629 630 632 The End-System Route Server will convert the information received in 633 a 'publish' request into the corresponding BGP route information such 634 that:. 636 It associates the specific request with a local VRF which it 637 resolves by using a combination of the originator jid and the 638 collection 'node' attribute. 640 It creates a BGP VPN route with a 'Route Distinguisher' (RD) which 641 contains a unique 32bit value per end-system plus a 16bit 642 instance-id, the specified IP prefix and 'label' received from the 643 VPN Forwarder as the Network Layer Reachability Information 644 (NLRI). The instance-id is either the value specified by the XMPP 645 client in the subscribe message for the specific pubsub node or a 646 locally generated value when that parameter is omitted. 648 The BGP next-hop address is set to the address of the VPN 649 Forwarder. 651 A BGP Tunnel Encapsulation Attribute [RFC5512] is generated for 652 each 'tunnel-encapsulation' element specified in the XMPP message. 654 It optionally associates the route with a MAC Mobility extended 655 community [I-D.ietf-l2vpn-evpn] containing a sequence number of 656 the route advertisement. 658 Conversely, when an interface operational status is determined to be 659 down or an IP address is unconfigured the VPN forwarder generates an 660 XMPP retract message to withdraw the route advertisement. 662 Retract request from VPN Forwarder to End-System Route Server: 664 668 669 670 671 672 673 674 Update notification from Route Server to VPN Forwarder: 676 677 678 679 680 681 682 1 683
'vpn-ip-address>/32'
684 685 686 687 1 688
'infrastructure-ip-address'
689 690 691 692 1 693 694 695 696 ... 697 698 699 700 702 Notifications should be generated whenever a VPN route is added, 703 modified or deleted. These notification messages contain only items 704 that have been added, modified or deleted since the previous 705 information sent to the VPN Forwarder. Notification messages can be 706 segmented at the convinience of the Router Server. 708 Note that the Update from the Route Server to the VPN Forwarder does 709 not contain the "jid" of the destination end-system. The "from" 710 attribute in the 'message' element contains a "jid" associated with 711 the Route Servers in the domain. The XMPP messages are point-to- 712 point in nature, between a Forwarder and Route Server. Even in the 713 case when one XMPP publish request from a Forwarder may cause the 714 Route Server to generate one or more event notifications. 716 The item "id" used in publish and retract messages must be unique 717 within the context of a XMPP pubsub node. This specification uses an 718 id format that corresponds to the string representation of the route 719 such that the leading part corresponds to an IP identifier of the 720 end-system, followed by the 'instance-id' for the specific VRF and 721 the IP prefix in its canonical string representation. 723 When multiple possible routes exist for a given VPN IP address within 724 a VRF it is the responsibility of the Route Server to select the best 725 path to advertise to the Forwarder. 727 A VPN Forwarder uses locally originated information to generate MPLS 728 label forwarding state, used to forward downstream traffic (i.e. 729 traffic received from the network). Upstream traffic (i.e. received 730 from a virtual-interface) is forwarded according to the routing 731 information received from one or more Route Servers that the VPN 732 forwarder has an XMPP session with. In the case where multiple 733 Router Servers are providing routing information for a specific NLRI 734 the VPN Forwarder SHOULD select the following algorithm: 736 Prefer the highest local-preference value; 738 Prefer the highest sequence-number; 740 Tie-break on the Route Server IP address. 742 When routes are withdrawn, the End-System Route Server generates an 743 item "retract" request. 745 Route advertisements can have an optional sequence-number which help 746 the route server determine the most recent route advertisement. The 747 sequence number is detemined by an mechanism external to this 748 document. One example is to use time synchronization between compute 749 nodes to have a globally coordinated timestamp. This timestamp can 750 be used to identify the time of interface creation on the compute 751 node. 753 Routes can also be associated with a "local-preference" attribute. 754 This attribute mapps to the BGP attribute of the same name for the 755 purposes of route selection. 757 7. End-System Route Server behavior 759 End-System Route Servers SHALL support the BGP address families: VPN- 760 IPv4 (1, 128), VPN-IPv6 (2, 128) and RT-Constraint (1, 132) 761 [RFC4684]. 763 When an End-System Route Server receives a request to create or 764 modify a VPN route it SHALL generate a BGP VPN route advertisement 765 with the corresponding information. 767 It is assumed that the End-System Route Servers have information 768 regarding the mapping between the tuple ('end-system', 'vpn-customer- 769 names') and BGP Route Targets used to import and export information 770 from the associated VRFs. This mapping is known via an out-of-band 771 mechanism not specified in this document. 773 Whenever the End-System Route Server receives an XMPP subscription 774 request, it SHALL consult its RT-Constraint Routing Information Base 775 (RIB). If the Route Server does not have a locally originated RT- 776 Constraint route that corresponds to the vpn-name present in the 777 request, it SHALL create one and generate the corresponding BGP route 778 advertisement. This route advertisement should only be withdrawn 779 when there are no more downstream XMPP clients subscribed to the VPN. 781 End-System Route Servers SHOULD automatically assign a BGP route 782 distinguisher per VPN routing table. 784 8. Operational Model 786 In the simplest case, a VPN is a collection of systems that are 787 allowed to exchange traffic with each other and only with each other. 788 Since all the forwarding tables in this VPN have the same routing 789 entries they are often referred to as symmetrical VPNs. 791 In order to better illustrate the operation of the protocol we 792 consider a simple example in which "host 1" and "host 2" both contain 793 a virtual interface that is a member of the same VPN. 795 Each of these hosts has an XMPP session with an End-System Route 796 Server, RS1 and RS2 our example, and these Route Servers are part of 797 the same BGP mesh. 799 When a virtual interface is created on "host 1", the local XMPP 800 client generates a XMPP subscription message to its respective Route 801 Server. This message contains a VPN identifier that has been 802 assigned by the provisioning system. The Route Server maps that 803 identifier to a BGP IP VPN configuration which contains the list of 804 import and export route targets to be used for that particular VRF. 806 Once the interface is operational, "host 1" will publish any IP 807 addresses that are configured on the respective virtual interface. 808 This will in turn cause the End-System Route Server to advertise 809 these (directly or indirectly) to any other BGP speaker on the 810 network which is connected to an attachment point of that VPN. 812 +--------+ +------------+ +----------+ 813 | host 1 | <===> | End-System | <===> | BGP mesh | 814 +--------+ |Route Server| +----------+ 815 +------------+ 817 Figure 1 819 +----------------+-------------+-------+-----------+ 820 | VPN IP address | NEXT-HOP | label | Known via | 821 +----------------+-------------+-------+-----------+ 822 | 10.1.1.1/32 | 192.168.1.1 | 10000 | XMPP | 823 | | | | | 824 | 10.1.1.2/32 | 192.168.2.1 | 20000 | BGP | 825 +----------------+-------------+-------+-----------+ 827 VPN Routing table on Route Server 829 Table 1 831 The figure above represents the contents of the VRF routing table on 832 RS1 after the IPv4 address 10.1.1.1 has been added to the virtual 833 interface on host 1. It assumes that there is another attachement 834 point for this VPN with the IPv4 address of 10.1.1.2. Host 1 has an 835 infrastructure IP address of 192.168.1.1 configured on its physical 836 interface while host 2 has IP address 192.168.2.1. 838 The contents of the VRF routing table in the End-System Route Servers 839 are advertised via XMPP Update notifications sent to host 1. This 840 information is the used by the host to populate the forwarding table 841 associated with that VPN. 843 +--------+ +--------+ 844 app -- veth0 --| host 1 |=== [network] ===| host 2 |-- veth0 -- app 845 +--------+ +--------+ 847 IP pkt ===> encap (GRE + label) ===> [IP net] ===> decap ===> IP pkt 848 [192.168.2.1, 20] map 20 to veth0 850 Figure 2 852 +----------------+--------------+-------+ 853 | VPN IP address | Host address | label | 854 +----------------+--------------+-------+ 855 | 10.1.1.1/32 | localhost | 10000 | 856 | | | | 857 | 10.1.1.2/32 | 192.168.2.1 | 20000 | 858 +----------------+--------------+-------+ 860 VRF table on host1 862 Table 2 864 When an application that uses the virtual interface on host 1 865 generates packets with a destination IP address of 10.1.1.2 these are 866 routed by the VPN Forwarder implemented in the Host OS. The packets 867 are encapsulated with a header that contains a 20-bit label assigned 868 by host 2. 870 In the case the virtual interface on host is associated with a guest 871 OS, this guest OS has had its address resolution queries answered 872 with the Virtual Router MAC address. As a result, that is the 873 address it uses as the destination MAC address in packets it 874 originates. This MAC address is not present on the encapsulated 875 packet. 877 End-System Route Servers are software applications that implement 878 both the BGP IP VPN PE control plane as well as XMPP server 879 functionality. These applications are not in the forwarding plane 880 and do not need to be co-located with a network device. 882 Network devices MAY have direct BGP sessions to the End-System Route 883 Servers. For instance, a router or security appliance that supports 884 BGP/MPLS IP VPNs over GRE may use its existing functionality to 885 inter-operate directly with a collection of Virtual Machines or other 886 network appliances that support this specification. 888 End-System Route Servers implement the VRF import policy and export 889 policy functionality that is associated with PE routers in standard 890 BGP IP/VPN deployments. VPN Forwarders receive forwarding 891 information after policy and route selection is applied. These are 892 unqualified routes in a specific VRF rather than VPN routing 893 information qualified by a Route Distinguisher and with a set of 894 Route Targets. 896 A symmetrical VPN uses a vrf import and vrf export polices that 897 contain a single route target, where the route target used for both 898 import and export is the same. 900 Different VPN topologies can be created by manipulating the vrf 901 import and export configuration including "hub-and-spoke" topologies 902 or overlapping VPNs. 904 An example of a hub-and-spoke VPN configuration is one where all the 905 traffic from the VPN clients must be redirected though a middle-box 906 for inspection. Assuming that the virtual interfaces of a particular 907 user are configured to be in the VPN "tenant1". At an initial stage 908 this "tenant1" VPN is symmetrical and uses a single Route Target in 909 both its import and export policies. The middle-box functionality 910 can be incrementally deployed by defining a new VPN, "tenant1-hub", 911 and an associated Route Target. Accompanied with a change in the 912 End-System Route Server configuration such that VPN "tenant1" only 913 imports routes with the Route Target associated with the hub. The 914 "hub" VPN is assumed to advertise a prefix that covers all the VPN 915 clients IP addresses. The "hub" VPN imports the VPN routes in order 916 for it to be able to generate the XMPP updates to the "hub" end- 917 system. This information is required for the return traffic from the 918 hub to the spokes (the VPN clients). In such a scenario a single 919 physical interface can connect the middle-box to the clients in a 920 given VPN which appear logically as downstream from it. Such a 921 middle-box would often require connectivity to multiple VPNs, such as 922 for instance an "outside" VPN which provides external connectivity to 923 one or more "inside" VPNs. 925 The functionality defined in this document in which the BGP IP VPN PE 926 functionality is split into its control (End-System Route Servers) 927 and forwarding (VPN Forwarder) components is fully interoperable with 928 existing BGP IP VPN PEs. 930 This makes it possible to reuse existing systems. For example, at 931 the edge of a data-center facility it may be desirable to use an 932 existing router or appliance that aggregates IP VPN routing 933 information and/or provides IP based services such as stateful packet 934 inspection. 936 Such a system can be configured, based on existing functionality, to 937 suppress more specific routes than a specified aggregate while 938 advertising the aggregate with a BGP NEXT_HOP containing the PE's IP 939 address and a locally assigned label corresponding to a VRF where the 940 more specific routes are present. 942 9. IANA Considerations 944 This document has no IANA actions. 946 10. Security Considerations 948 The signaling protocol defines the access control policies for each 949 virtual interface and any guest application associated with it. It 950 is important to secure the end-system access to End-System Route 951 Servers and the BGP infrastructure itself. 953 The XMPP session between end-systems and the Route Servers MUST use 954 mutual authentication. One possible strategy is to distribute pre- 955 signed certificates to end-systems which are presented as proof of 956 authorization to the Route Server. 958 BGP sessions MUST be authenticated. This document recommends that 959 BGP speaking systems filter traffic on port 179 such that only IP 960 addresses which are known to participate in the BGP signaling 961 protocol are allowed. 963 As a security measure, it is recommended that virtual and 964 infrastructure topologies never be allowed to exchange traffic 965 directly. The infrastructure network containing the end-systems is 966 typically isolated from the outside world. 968 11. XML schema 970 The following schema defines the XML elements that are used to 971 communicate unicast reachability information between the Route Server 972 and VPN Forwarder: 974 978 979 980 981 982 983 984 986 987 988 991 992 994 995 996 997 998 999 1001 1002 1004 1005 1006 1008 1009 1010 1011 1012 1013 1014 1015 1016 1018 1019 1020 1021 1022 1023 1024 1025 1027 1028 1029 1030 1031 1033 1034 1035 1037 1038 1040 1042 1044 12. Acknowledgements 1046 Yakov Rekhter has contributed to this document by providing detailed 1047 feedback and suggestions. The authors would also like to thank 1048 Thomas Morin for his comments. 1050 Amit Shukla and Ping Pan contributed to earlier versions of this 1051 document. 1053 13. References 1054 13.1. Normative References 1056 [RFC1027] Carl-Mitchell, S. and J. Quarterman, "Using ARP to 1057 implement transparent subnet gateways", RFC 1027, October 1058 1987. 1060 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1061 Requirement Levels", BCP 14, RFC 2119, March 1997. 1063 [RFC4023] Worster, T., Rekhter, Y., and E. Rosen, "Encapsulating 1064 MPLS in IP or Generic Routing Encapsulation (GRE)", RFC 1065 4023, March 2005. 1067 [RFC4364] Rosen, E. and Y. Rekhter, "BGP/MPLS IP Virtual Private 1068 Networks (VPNs)", RFC 4364, February 2006. 1070 [RFC4456] Bates, T., Chen, E., and R. Chandra, "BGP Route 1071 Reflection: An Alternative to Full Mesh Internal BGP 1072 (IBGP)", RFC 4456, April 2006. 1074 [RFC4684] Marques, P., Bonica, R., Fang, L., Martini, L., Raszuk, 1075 R., Patel, K., and J. Guichard, "Constrained Route 1076 Distribution for Border Gateway Protocol/MultiProtocol 1077 Label Switching (BGP/MPLS) Internet Protocol (IP) Virtual 1078 Private Networks (VPNs)", RFC 4684, November 2006. 1080 [RFC5512] Mohapatra, P. and E. Rosen, "The BGP Encapsulation 1081 Subsequent Address Family Identifier (SAFI) and the BGP 1082 Tunnel Encapsulation Attribute", RFC 5512, April 2009. 1084 [RFC5798] Nadas, S., "Virtual Router Redundancy Protocol (VRRP) 1085 Version 3 for IPv4 and IPv6", RFC 5798, March 2010. 1087 [RFC6120] Saint-Andre, P., "Extensible Messaging and Presence 1088 Protocol (XMPP): Core", RFC 6120, March 2011. 1090 [pubsub] Millard, P., Saint-Andre, P., and R. Meijer, "Publish- 1091 Subscribe", XEP 0060, July 2010. 1093 13.2. Informational References 1095 [RFC5575] Marques, P., Sheth, N., Raszuk, R., Greene, B., Mauch, J., 1096 and D. McPherson, "Dissemination of Flow Specification 1097 Rules", RFC 5575, August 2009. 1099 [I-D.ietf-mpls-in-udp] 1100 Xu, X., Sheth, N., Yong, L., Pignataro, C., and F. 1101 Yongbing, "Encapsulating MPLS in UDP", draft-ietf-mpls-in- 1102 udp-05 (work in progress), January 2014. 1104 [I-D.ietf-l2vpn-evpn] 1105 Sajassi, A., Aggarwal, R., Bitar, N., Isaac, A., and J. 1106 Uttaro, "BGP MPLS Based Ethernet VPN", draft-ietf-l2vpn- 1107 evpn-08 (work in progress), September 2014. 1109 [IEEE.802-1Q] 1110 Institute of Electrical and Electronics Engineers, "Local 1111 and Metropolitan Area Networks: Virtual Bridged Local Area 1112 Networks", IEEE Std 802.1Q-2005, May 2006. 1114 Authors' Addresses 1116 Pedro Marques 1117 Juniper Networks 1118 1133 Innovation Way 1119 Sunnyvale, CA 94089 1121 Email: roque@juniper.net 1123 Luyuan Fang 1124 Microsoft 1125 5600 148th Ave NE 1126 Redmond, WA 98052 1128 Email: lufang@microsoft.com 1130 Nischal Sheth 1131 Juniper Networks 1132 1133 Innovation Way 1133 Sunnyvale, CA 94089 1135 Email: nsheth@juniper.net 1137 Maria Napierala 1138 AT&T Labs 1139 200 Laurel Avenue 1140 Middletown, NJ 07748 1142 Email: mnapierala@att.com 1143 Nabil Bitar 1144 Verizon 1145 40 Sylvan Rd. 1146 Waltham, MA 02145 1148 Email: nabil.bitar@verizon.com