idnits 2.17.1 draft-ietf-nvo3-geneve-03.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (September 20, 2016) is 2768 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) ** Obsolete normative reference: RFC 5226 (Obsoleted by RFC 8126) -- Obsolete informational reference (is this intentional?): RFC 1981 (Obsoleted by RFC 8201) Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Gross, Ed. 3 Internet-Draft 4 Intended status: Standards Track I. Ganga, Ed. 5 Expires: March 24, 2017 Intel 6 T. Sridhar, Ed. 7 VMware 8 September 20, 2016 10 Geneve: Generic Network Virtualization Encapsulation 11 draft-ietf-nvo3-geneve-03 13 Abstract 15 Network virtualization involves the cooperation of devices with a 16 wide variety of capabilities such as software and hardware tunnel 17 endpoints, transit fabrics, and centralized control clusters. As a 18 result of their role in tying together different elements in the 19 system, the requirements on tunnels are influenced by all of these 20 components. Flexibility is therefore the most important aspect of a 21 tunnel protocol if it is to keep pace with the evolution of the 22 system. This draft describes Geneve, a protocol designed to 23 recognize and accommodate these changing capabilities and needs. 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 March 24, 2017. 42 Copyright Notice 44 Copyright (c) 2016 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 . . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 4 61 1.2. Terminology . . . . . . . . . . . . . . . . . . . . . . . 4 62 2. Design Requirements . . . . . . . . . . . . . . . . . . . . . 5 63 2.1. Control Plane Independence . . . . . . . . . . . . . . . 6 64 2.2. Data Plane Extensibility . . . . . . . . . . . . . . . . 7 65 2.2.1. Efficient Implementation . . . . . . . . . . . . . . 7 66 2.3. Use of Standard IP Fabrics . . . . . . . . . . . . . . . 8 67 3. Geneve Encapsulation Details . . . . . . . . . . . . . . . . 9 68 3.1. Geneve Packet Format Over IPv4 . . . . . . . . . . . . . 9 69 3.2. Geneve Packet Format Over IPv6 . . . . . . . . . . . . . 10 70 3.3. UDP Header . . . . . . . . . . . . . . . . . . . . . . . 12 71 3.4. Tunnel Header Fields . . . . . . . . . . . . . . . . . . 13 72 3.5. Tunnel Options . . . . . . . . . . . . . . . . . . . . . 14 73 3.5.1. Options Processing . . . . . . . . . . . . . . . . . 16 74 4. Implementation and Deployment Considerations . . . . . . . . 17 75 4.1. Encapsulation of Geneve in IP . . . . . . . . . . . . . . 17 76 4.1.1. IP Fragmentation . . . . . . . . . . . . . . . . . . 17 77 4.1.2. DSCP and ECN . . . . . . . . . . . . . . . . . . . . 17 78 4.1.3. Broadcast and Multicast . . . . . . . . . . . . . . . 18 79 4.1.4. Unidirectional Tunnels . . . . . . . . . . . . . . . 18 80 4.2. Constraints on Protocol Features . . . . . . . . . . . . 19 81 4.3. NIC Offloads . . . . . . . . . . . . . . . . . . . . . . 19 82 4.4. Inner VLAN Handling . . . . . . . . . . . . . . . . . . . 20 83 5. Interoperability Issues . . . . . . . . . . . . . . . . . . . 20 84 6. Security Considerations . . . . . . . . . . . . . . . . . . . 21 85 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 86 8. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 22 87 9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 23 88 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 23 89 10.1. Normative References . . . . . . . . . . . . . . . . . . 23 90 10.2. Informative References . . . . . . . . . . . . . . . . . 24 91 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 25 93 1. Introduction 95 Networking has long featured a variety of tunneling, tagging, and 96 other encapsulation mechanisms. However, the advent of network 97 virtualization has caused a surge of renewed interest and a 98 corresponding increase in the introduction of new protocols. The 99 large number of protocols in this space, ranging all the way from 100 VLANs [IEEE.802.1Q-2014] and MPLS [RFC3031] through the more recent 101 VXLAN [RFC7348], NVGRE [RFC7637], and STT [I-D.davie-stt], often 102 leads to questions about the need for new encapsulation formats and 103 what it is about network virtualization in particular that leads to 104 their proliferation. 106 While many encapsulation protocols seek to simply partition the 107 underlay network or bridge between two domains, network 108 virtualization views the transit network as providing connectivity 109 between multiple components of a distributed system. In many ways 110 this system is similar to a chassis switch with the IP underlay 111 network playing the role of the backplane and tunnel endpoints on the 112 edge as line cards. When viewed in this light, the requirements 113 placed on the tunnel protocol are significantly different in terms of 114 the quantity of metadata necessary and the role of transit nodes. 116 Current work such as [VL2] and the NVO3 working group 117 [I-D.ietf-nvo3-dataplane-requirements] have described some of the 118 properties that the data plane must have to support network 119 virtualization. However, one additional defining requirement is the 120 need to carry system state along with the packet data. The use of 121 some metadata is certainly not a foreign concept - nearly all 122 protocols used for virtualization have at least 24 bits of identifier 123 space as a way to partition between tenants. This is often described 124 as overcoming the limits of 12-bit VLANs, and when seen in that 125 context, or any context where it is a true tenant identifier, 16 126 million possible entries is a large number. However, the reality is 127 that the metadata is not exclusively used to identify tenants and 128 encoding other information quickly starts to crowd the space. In 129 fact, when compared to the tags used to exchange metadata between 130 line cards on a chassis switch, 24-bit identifiers start to look 131 quite small. There are nearly endless uses for this metadata, 132 ranging from storing input ports for simple security policies to 133 service based context for interposing advanced middleboxes. 135 Existing tunnel protocols have each attempted to solve different 136 aspects of these new requirements, only to be quickly rendered out of 137 date by changing control plane implementations and advancements. 138 Furthermore, software and hardware components and controllers all 139 have different advantages and rates of evolution - a fact that should 140 be viewed as a benefit, not a liability or limitation. This draft 141 describes Geneve, a protocol which seeks to avoid these problems by 142 providing a framework for tunneling for network virtualization rather 143 than being prescriptive about the entire system. 145 1.1. Requirements Language 147 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 148 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 149 document are to be interpreted as described in [RFC2119]. 151 In this document, these words will appear with that interpretation 152 only when in ALL CAPS. Lower case uses of these words are not to be 153 interpreted as carrying RFC-2119 significance. 155 1.2. Terminology 157 The NVO3 framework [RFC7365] defines many of the concepts commonly 158 used in network virtualization. In addition, the following terms are 159 specifically meaningful in this document: 161 Checksum offload. An optimization implemented by many NICs which 162 enables computation and verification of upper layer protocol 163 checksums in hardware on transmit and receive, respectively. This 164 typically includes IP and TCP/UDP checksums which would otherwise be 165 computed by the protocol stack in software. 167 Clos network. A technique for composing network fabrics larger than 168 a single switch while maintaining non-blocking bandwidth across 169 connection points. ECMP is used to divide traffic across the 170 multiple links and switches that constitute the fabric. Sometimes 171 termed "leaf and spine" or "fat tree" topologies. 173 ECMP. Equal Cost Multipath. A routing mechanism for selecting from 174 among multiple best next hop paths by hashing packet headers in order 175 to better utilize network bandwidth while avoiding reordering a 176 single stream. 178 Geneve. Generic Network Virtualization Encapsulation. The tunnel 179 protocol described in this draft. 181 LRO. Large Receive Offload. The receive-side equivalent function of 182 LSO, in which multiple protocol segments (primarily TCP) are 183 coalesced into larger data units. 185 NIC. Network Interface Card. A NIC could be part of a tunnel 186 endpoint or transit device and can either process Geneve packets or 187 aid in the processing of Geneve packets. 189 OAM. Operations, Administration, and Management. A suite of tools 190 used to monitor and troubleshoot network problems. 192 Transit device. A forwarding element along the path of the tunnel 193 making up part of the Underlay Network. A transit device MAY be 194 capable of understanding the Geneve packet format but does not 195 originate or terminate Geneve packets. 197 LSO. Large Segmentation Offload. A function provided by many 198 commercial NICs that allows data units larger than the MTU to be 199 passed to the NIC to improve performance, the NIC being responsible 200 for creating smaller segments of size less than or equal to the MTU 201 with correct protocol headers. When referring specifically to TCP/ 202 IP, this feature is often known as TSO (TCP Segmentation Offload). 204 Tunnel endpoint. A component performing encapsulation and 205 decapsulation of packets, such as Ethernet frames or IP datagrams, in 206 Geneve headers. As the ultimate consumer of any tunnel metadata, 207 endpoints have the highest level of requirements for parsing and 208 interpreting tunnel headers. Tunnel endpoints may consist of either 209 software or hardware implementations or a combination of the two. 210 Endpoints are frequently a component of an NVE but may also be found 211 in middleboxes or other elements making up an NVO3 Network. 213 VM. Virtual Machine. 215 2. Design Requirements 217 Geneve is designed to support network virtualization use cases, where 218 tunnels are typically established to act as a backplane between the 219 virtual switches residing in hypervisors, physical switches, or 220 middleboxes or other appliances. An arbitrary IP network can be used 221 as an underlay although Clos networks composed using ECMP links are a 222 common choice to provide consistent bisectional bandwidth across all 223 connection points. Figure 1 shows an example of a hypervisor, top of 224 rack switch for connectivity to physical servers, and a WAN uplink 225 connected using Geneve tunnels over a simplified Clos network. These 226 tunnels are used to encapsulate and forward frames from the attached 227 components such as VMs or physical links. 229 +---------------------+ +-------+ +------+ 230 | +--+ +-------+---+ | |Transit|--|Top of|==Physical 231 | |VM|--| | | | +------+ /|Router | | Rack |==Servers 232 | +--+ |Virtual|NIC|---|Top of|/ +-------+\/+------+ 233 | +--+ |Switch | | | | Rack |\ +-------+/\+------+ 234 | |VM|--| | | | +------+ \|Transit| |Uplink| WAN 235 | +--+ +-------+---+ | |Router |--| |=========> 236 +---------------------+ +-------+ +------+ 237 Hypervisor 239 ()===================================() 240 Switch-Switch Geneve Tunnels 242 Figure 1: Sample Geneve Deployment 244 To support the needs of network virtualization, the tunnel protocol 245 should be able to take advantage of the differing (and evolving) 246 capabilities of each type of device in both the underlay and overlay 247 networks. This results in the following requirements being placed on 248 the data plane tunneling protocol: 250 o The data plane is generic and extensible enough to support current 251 and future control planes. 253 o Tunnel components are efficiently implementable in both hardware 254 and software without restricting capabilities to the lowest common 255 denominator. 257 o High performance over existing IP fabrics. 259 These requirements are described further in the following 260 subsections. 262 2.1. Control Plane Independence 264 Although some protocols for network virtualization have included a 265 control plane as part of the tunnel format specification (most 266 notably, the original VXLAN spec prescribed a multicast learning- 267 based control plane), these specifications have largely been treated 268 as describing only the data format. The VXLAN packet format has 269 actually seen a wide variety of control planes built on top of it. 271 There is a clear advantage in settling on a data format: most of the 272 protocols are only superficially different and there is little 273 advantage in duplicating effort. However, the same cannot be said of 274 control planes, which are diverse in very fundamental ways. The case 275 for standardization is also less clear given the wide variety in 276 requirements, goals, and deployment scenarios. 278 As a result of this reality, Geneve aims to be a pure tunnel format 279 specification that is capable of fulfilling the needs of many control 280 planes by explicitly not selecting any one of them. This 281 simultaneously promotes a shared data format and increases the 282 chances that it will not be obsoleted by future control plane 283 enhancements. 285 2.2. Data Plane Extensibility 287 Achieving the level of flexibility needed to support current and 288 future control planes effectively requires an options infrastructure 289 to allow new metadata types to be defined, deployed, and either 290 finalized or retired. Options also allow for differentiation of 291 products by encouraging independent development in each vendor's core 292 specialty, leading to an overall faster pace of advancement. By far 293 the most common mechanism for implementing options is Type-Length- 294 Value (TLV) format. 296 It should be noted that while options can be used to support non- 297 wirespeed control packets, they are equally important on data packets 298 as well to segregate and direct forwarding (for instance, the 299 examples given before of input port based security policies and 300 service interposition both require tags to be placed on data 301 packets). Therefore, while it would be desirable to limit the 302 extensibility to only control packets for the purposes of simplifying 303 the datapath, that would not satisfy the design requirements. 305 2.2.1. Efficient Implementation 307 There is often a conflict between software flexibility and hardware 308 performance that is difficult to resolve. For a given set of 309 functionality, it is obviously desirable to maximize performance. 310 However, that does not mean new features that cannot be run at that 311 speed today should be disallowed. Therefore, for a protocol to be 312 efficiently implementable means that a set of common capabilities can 313 be reasonably handled across platforms along with a graceful 314 mechanism to handle more advanced features in the appropriate 315 situations. 317 The use of a variable length header and options in a protocol often 318 raises questions about whether it is truly efficiently implementable 319 in hardware. To answer this question in the context of Geneve, it is 320 important to first divide "hardware" into two categories: tunnel 321 endpoints and transit devices. 323 Endpoints must be able to parse the variable header, including any 324 options, and take action. Since these devices are actively 325 participating in the protocol, they are the most affected by Geneve. 327 However, as endpoints are the ultimate consumers of the data, 328 transmitters can tailor their output to the capabilities of the 329 recipient. As new functionality becomes sufficiently well defined to 330 add to endpoints, supporting options can be designed using ordering 331 restrictions and other techniques to ease parsing. 333 Transit devices MAY be able to interpret the options and participate 334 in Geneve packet processing. However, as non-terminating devices, 335 they do not originate or terminate the Geneve packet. The 336 participation of transit devices in Geneve packet processing is 337 OPTIONAL. 339 Further, either tunnel endpoints or transit devices MAY use offload 340 capabilities of NICs such as checksum offload to improve the 341 performance of Geneve packet processing. The presence of a Geneve 342 variable length header SHOULD NOT prevent the tunnel endpoints and 343 transit devices from using such offload capabilities. 345 2.3. Use of Standard IP Fabrics 347 IP has clearly cemented its place as the dominant transport mechanism 348 and many techniques have evolved over time to make it robust, 349 efficient, and inexpensive. As a result, it is natural to use IP 350 fabrics as a transit network for Geneve. Fortunately, the use of IP 351 encapsulation and addressing is enough to achieve the primary goal of 352 delivering packets to the correct point in the network through 353 standard switching and routing. 355 In addition, nearly all underlay fabrics are designed to exploit 356 parallelism in traffic to spread load across multiple links without 357 introducing reordering in individual flows. These equal cost 358 multipathing (ECMP) techniques typically involve parsing and hashing 359 the addresses and port numbers from the packet to select an outgoing 360 link. However, the use of tunnels often results in poor ECMP 361 performance without additional knowledge of the protocol as the 362 encapsulated traffic is hidden from the fabric by design and only 363 endpoint addresses are available for hashing. 365 Since it is desirable for Geneve to perform well on these existing 366 fabrics, it is necessary for entropy from encapsulated packets to be 367 exposed in the tunnel header. The most common technique for this is 368 to use the UDP source port, which is discussed further in 369 Section 3.3. 371 3. Geneve Encapsulation Details 373 The Geneve packet format consists of a compact tunnel header 374 encapsulated in UDP over either IPv4 or IPv6. A small fixed tunnel 375 header provides control information plus a base level of 376 functionality and interoperability with a focus on simplicity. This 377 header is then followed by a set of variable options to allow for 378 future innovation. Finally, the payload consists of a protocol data 379 unit of the indicated type, such as an Ethernet frame. Section 3.1 380 and Section 3.2 illustrate the Geneve packet format transported (for 381 example) over Ethernet along with an Ethernet payload. 383 3.1. Geneve Packet Format Over IPv4 385 0 1 2 3 386 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 388 Outer Ethernet Header: 389 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 390 | Outer Destination MAC Address | 391 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 392 | Outer Destination MAC Address | Outer Source MAC Address | 393 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 394 | Outer Source MAC Address | 395 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 396 |Optional Ethertype=C-Tag 802.1Q| Outer VLAN Tag Information | 397 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 398 | Ethertype=0x0800 | 399 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 401 Outer IPv4 Header: 402 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 403 |Version| IHL |Type of Service| Total Length | 404 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 405 | Identification |Flags| Fragment Offset | 406 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 407 | Time to Live |Protocol=17 UDP| Header Checksum | 408 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 409 | Outer Source IPv4 Address | 410 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 411 | Outer Destination IPv4 Address | 412 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 414 Outer UDP Header: 415 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 416 | Source Port = xxxx | Dest Port = 6081 | 417 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 418 | UDP Length | UDP Checksum | 419 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 421 Geneve Header: 422 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 423 |Ver| Opt Len |O|C| Rsvd. | Protocol Type | 424 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 425 | Virtual Network Identifier (VNI) | Reserved | 426 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 427 | Variable Length Options | 428 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 430 Inner Ethernet Header (example payload): 431 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 432 | Inner Destination MAC Address | 433 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 434 | Inner Destination MAC Address | Inner Source MAC Address | 435 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 436 | Inner Source MAC Address | 437 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 438 |Optional Ethertype=C-Tag 802.1Q| Inner VLAN Tag Information | 439 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 441 Payload: 442 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 443 | Ethertype of Original Payload | | 444 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | 445 | Original Ethernet Payload | 446 | | 447 | (Note that the original Ethernet Frame's FCS is not included) | 448 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 450 Frame Check Sequence: 451 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 452 | New FCS (Frame Check Sequence) for Outer Ethernet Frame | 453 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 455 3.2. Geneve Packet Format Over IPv6 456 0 1 2 3 457 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 459 Outer Ethernet Header: 460 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 461 | Outer Destination MAC Address | 462 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 463 | Outer Destination MAC Address | Outer Source MAC Address | 464 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 465 | Outer Source MAC Address | 466 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 467 |Optional Ethertype=C-Tag 802.1Q| Outer VLAN Tag Information | 468 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 469 | Ethertype=0x86DD | 470 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 472 Outer IPv6 Header: 473 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 474 |Version| Traffic Class | Flow Label | 475 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 476 | Payload Length | NxtHdr=17 UDP | Hop Limit | 477 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 478 | | 479 + + 480 | | 481 + Outer Source IPv6 Address + 482 | | 483 + + 484 | | 485 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 486 | | 487 + + 488 | | 489 + Outer Destination IPv6 Address + 490 | | 491 + + 492 | | 493 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 495 Outer UDP Header: 496 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 497 | Source Port = xxxx | Dest Port = 6081 | 498 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 499 | UDP Length | UDP Checksum | 500 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 502 Geneve Header: 503 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 504 |Ver| Opt Len |O|C| Rsvd. | Protocol Type | 505 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 506 | Virtual Network Identifier (VNI) | Reserved | 507 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 508 | Variable Length Options | 509 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 511 Inner Ethernet Header (example payload): 512 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 513 | Inner Destination MAC Address | 514 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 515 | Inner Destination MAC Address | Inner Source MAC Address | 516 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 517 | Inner Source MAC Address | 518 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 519 |Optional Ethertype=C-Tag 802.1Q| Inner VLAN Tag Information | 520 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 522 Payload: 523 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 524 | Ethertype of Original Payload | | 525 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | 526 | Original Ethernet Payload | 527 | | 528 | (Note that the original Ethernet Frame's FCS is not included) | 529 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 531 Frame Check Sequence: 532 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 533 | New FCS (Frame Check Sequence) for Outer Ethernet Frame | 534 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 536 3.3. UDP Header 538 The use of an encapsulating UDP [RFC0768] header follows the 539 connectionless semantics of Ethernet and IP in addition to providing 540 entropy to routers performing ECMP. The header fields are therefore 541 interpreted as follows: 543 Source port: A source port selected by the originating tunnel 544 endpoint. This source port SHOULD be the same for all packets 545 belonging to a single encapsulated flow to prevent reordering due 546 to the use of different paths. To encourage an even distribution 547 of flows across multiple links, the source port SHOULD be 548 calculated using a hash of the encapsulated packet headers using, 549 for example, a traditional 5-tuple. Since the port represents a 550 flow identifier rather than a true UDP connection, the entire 551 16-bit range MAY be used to maximize entropy. 553 Dest port: IANA has assigned port 6081 as the fixed well-known 554 destination port for Geneve. Although the well-known value should 555 be used by default, it is RECOMMENDED that implementations make 556 this configurable. The chosen port is used for identification of 557 Geneve packets and MUST NOT be reversed for different ends of a 558 connection as is done with TCP. 560 UDP length: The length of the UDP packet including the UDP header. 562 UDP checksum: The checksum MAY be set to zero on transmit for 563 packets encapsulated in both IPv4 and IPv6 [RFC6935]. When a 564 packet is received with a UDP checksum of zero it MUST be accepted 565 and decapsulated. If the originating tunnel endpoint optionally 566 encapsulates a packet with a non-zero checksum, it MUST be a 567 correctly computed UDP checksum. Upon receiving such a packet, 568 the egress endpoint MUST validate the checksum. If the checksum 569 is not correct, the packet MUST be dropped, otherwise the packet 570 MUST be accepted for decapsulation. It is RECOMMENDED that the 571 UDP checksum be computed to protect the Geneve header and options 572 in situations where the network reliability is not high and the 573 packet is not protected by another checksum or CRC. 575 3.4. Tunnel Header Fields 577 Ver (2 bits): The current version number is 0. Packets received by 578 an endpoint with an unknown version MUST be dropped. Non- 579 terminating devices processing Geneve packets with an unknown 580 version number MUST treat them as UDP packets with an unknown 581 payload. 583 Opt Len (6 bits): The length of the options fields, expressed in 584 four byte multiples, not including the eight byte fixed tunnel 585 header. This results in a minimum total Geneve header size of 8 586 bytes and a maximum of 260 bytes. The start of the payload 587 headers can be found using this offset from the end of the base 588 Geneve header. 590 O (1 bit): OAM packet. This packet contains a control message 591 instead of a data payload. Endpoints MUST NOT forward the payload 592 and transit devices MUST NOT attempt to interpret or process it. 593 Since these are infrequent control messages, it is RECOMMENDED 594 that endpoints direct these packets to a high priority control 595 queue (for example, to direct the packet to a general purpose CPU 596 from a forwarding ASIC or to separate out control traffic on a 597 NIC). Transit devices MUST NOT alter forwarding behavior on the 598 basis of this bit, such as ECMP link selection. 600 C (1 bit): Critical options present. One or more options has the 601 critical bit set (see Section 3.5). If this bit is set then 602 tunnel endpoints MUST parse the options list to interpret any 603 critical options. On endpoints where option parsing is not 604 supported the packet MUST be dropped on the basis of the 'C' bit 605 in the base header. If the bit is not set tunnel endpoints MAY 606 strip all options using 'Opt Len' and forward the decapsulated 607 packet. Transit devices MUST NOT drop or modify packets on the 608 basis of this bit. 610 Rsvd. (6 bits): Reserved field which MUST be zero on transmission 611 and ignored on receipt. 613 Protocol Type (16 bits): The type of the protocol data unit 614 appearing after the Geneve header. This follows the EtherType 615 [ETYPES] convention with Ethernet itself being represented by the 616 value 0x6558. 618 Virtual Network Identifier (VNI) (24 bits): An identifier for a 619 unique element of a virtual network. In many situations this may 620 represent an L2 segment, however, the control plane defines the 621 forwarding semantics of decapsulated packets. The VNI MAY be used 622 as part of ECMP forwarding decisions or MAY be used as a mechanism 623 to distinguish between overlapping address spaces contained in the 624 encapsulated packet when load balancing across CPUs. 626 Reserved (8 bits): Reserved field which MUST be zero on transmission 627 and ignored on receipt. 629 Transit devices MUST maintain consistent forwarding behavior 630 irrespective of the value of 'Opt Len', including ECMP link 631 selection. These devices SHOULD be able to forward packets 632 containing options without resorting to a slow path. 634 3.5. Tunnel Options 636 0 1 2 3 637 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 638 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 639 | Option Class | Type |R|R|R| Length | 640 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 641 | Variable Option Data | 642 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 644 Geneve Option 646 The base Geneve header is followed by zero or more options in Type- 647 Length-Value format. Each option consists of a four byte option 648 header and a variable amount of option data interpreted according to 649 the type. 651 Option Class (16 bits): Namespace for the 'Type' field. IANA will 652 be requested to create a "Geneve Option Class" registry to 653 allocate identifiers for organizations, technologies, and vendors 654 that have an interest in creating types for options. Each 655 organization may allocate types independently to allow 656 experimentation and rapid innovation. It is expected that over 657 time certain options will become well known and a given 658 implementation may use option types from a variety of sources. In 659 addition, IANA will be requested to reserve specific ranges for 660 standardized and experimental options. 662 Type (8 bits): Type indicating the format of the data contained in 663 this option. Options are primarily designed to encourage future 664 extensibility and innovation and so standardized forms of these 665 options will be defined in a separate document. 667 The high order bit of the option type indicates that this is a 668 critical option. If the receiving endpoint does not recognize 669 this option and this bit is set then the packet MUST be dropped. 670 If the critical bit is set in any option then the 'C' bit in the 671 Geneve base header MUST also be set. Transit devices MUST NOT 672 drop packets on the basis of this bit. The following figure shows 673 the location of the 'C' bit in the 'Type' field: 675 0 1 2 3 4 5 6 7 8 676 +-+-+-+-+-+-+-+-+ 677 |C| Type | 678 +-+-+-+-+-+-+-+-+ 680 The requirement to drop a packet with an unknown critical option 681 applies to the entire tunnel endpoint system and not a particular 682 component of the implementation. For example, in a system 683 comprised of a forwarding ASIC and a general purpose CPU, this 684 does not mean that the packet must be dropped in the ASIC. An 685 implementation may send the packet to the CPU using a rate-limited 686 control channel for slow-path exception handling. 688 R (3 bits): Option control flags reserved for future use. MUST be 689 zero on transmission and ignored on receipt. 691 Length (5 bits): Length of the option, expressed in four byte 692 multiples excluding the option header. The total length of each 693 option may be between 4 and 128 bytes. Packets in which the total 694 length of all options is not equal to the 'Opt Len' in the base 695 header are invalid and MUST be silently dropped if received by an 696 endpoint. 698 Variable Option Data: Option data interpreted according to 'Type'. 700 3.5.1. Options Processing 702 Geneve options are primarily intended to be originated and processed 703 by tunnel endpoints. However, options MAY be processed by transit 704 devices along the tunnel path as well. Transit devices not 705 processing Geneve headers SHOULD process Geneve packets as any other 706 UDP packet and maintain consistent forwarding behavior. 708 In tunnel endpoints, the generation and interpretation of options is 709 determined by the control plane, which is out of the scope of this 710 document. However, to ensure interoperability between heterogeneous 711 devices some requirements are imposed on options and the devices that 712 process them: 714 o Receiving endpoints MUST drop packets containing unknown options 715 with the 'C' bit set in the option type. Conversely, transit 716 devices MUST NOT drop packets as a result of encountering unknown 717 options, including those with the 'C' bit set. 719 o Some options may be defined in such a way that the position in the 720 option list is significant. Therefore, options MUST NOT be 721 reordered by transit devices. 723 o An option MUST NOT affect the parsing or interpretation of any 724 other option. 726 When designing a Geneve option, it is important to consider how the 727 option will evolve in the future. Once an option is defined it is 728 reasonable to expect that implementations may come to depend on a 729 specific behavior. As a result, the scope of any future changes must 730 be carefully described upfront. 732 Unexpectedly significant interoperability issues may result from 733 changing the length of an option that was defined to be a certain 734 size. A particular option is specified to have either a fixed 735 length, which is constant, or a variable length, which may change 736 over time or for different use cases. This property is part of the 737 definition of the option and conveyed by the 'Type'. For fixed 738 length options, some implementations may choose to ignore the length 739 field in the option header and instead parse based on the well known 740 length associated with the type. In this case, redefining the length 741 will impact not only parsing of the option in question but also any 742 options that follow. Therefore, options that are defined to be fixed 743 length in size MUST NOT be redefined to a different length. Instead, 744 a new 'Type' should be allocated. 746 4. Implementation and Deployment Considerations 748 4.1. Encapsulation of Geneve in IP 750 As an IP-based tunnel protocol, Geneve shares many properties and 751 techniques with existing protocols. The application of some of these 752 are described in further detail, although in general most concepts 753 applicable to the IP layer or to IP tunnels generally also function 754 in the context of Geneve. 756 4.1.1. IP Fragmentation 758 To prevent fragmentation and maximize performance, the best practice 759 when using Geneve is to ensure that the MTU of the physical network 760 is greater than or equal to the MTU of the encapsulated network plus 761 tunnel headers. Manual or upper layer (such as TCP MSS clamping) 762 configuration can be used to ensure that fragmentation never takes 763 place, however, in some situations this may not be feasible. 765 It is strongly RECOMMENDED that Path MTU Discovery ([RFC1191], 766 [RFC1981]) be used by setting the DF bit in the IP header when Geneve 767 packets are transmitted over IPv4 (this is the default with IPv6). 768 The use of Path MTU Discovery on the transit network provides the 769 encapsulating endpoint with soft-state about the link that it may use 770 to prevent or minimize fragmentation depending on its role in the 771 virtualized network. 773 Note that some implementations may not be capable of supporting 774 fragmentation or other less common features of the IP header, such as 775 options and extension headers. 777 4.1.2. DSCP and ECN 779 When encapsulating IP (including over Ethernet) packets in Geneve, 780 there are several considerations for propagating DSCP and ECN bits 781 from the inner header to the tunnel on transmission and the reverse 782 on reception. 784 [RFC2983] provides guidance for mapping DSCP between inner and outer 785 IP headers. Network virtualization is typically more closely aligned 786 with the Pipe model described, where the DSCP value on the tunnel 787 header is set based on a policy (which may be a fixed value, one 788 based on the inner traffic class, or some other mechanism for 789 grouping traffic). Aspects of the Uniform model (which treats the 790 inner and outer DSCP value as a single field by copying on ingress 791 and egress) may also apply, such as the ability to remark the inner 792 header on tunnel egress based on transit marking. However, the 793 Uniform model is not conceptually consistent with network 794 virtualization, which seeks to provide strong isolation between 795 encapsulated traffic and the physical network. 797 [RFC6040] describes the mechanism for exposing ECN capabilities on IP 798 tunnels and propagating congestion markers to the inner packets. 799 This behavior MUST be followed for IP packets encapsulated in Geneve. 801 4.1.3. Broadcast and Multicast 803 Geneve tunnels may either be point-to-point unicast between two 804 endpoints or may utilize broadcast or multicast addressing. It is 805 not required that inner and outer addressing match in this respect. 806 For example, in physical networks that do not support multicast, 807 encapsulated multicast traffic may be replicated into multiple 808 unicast tunnels or forwarded by policy to a unicast location 809 (possibly to be replicated there). 811 With physical networks that do support multicast it may be desirable 812 to use this capability to take advantage of hardware replication for 813 encapsulated packets. In this case, multicast addresses may be 814 allocated in the physical network corresponding to tenants, 815 encapsulated multicast groups, or some other factor. The allocation 816 of these groups is a component of the control plane and therefore 817 outside of the scope of this document. When physical multicast is in 818 use, the 'C' bit in the Geneve header may be used with groups of 819 devices with heterogeneous capabilities as each device can interpret 820 only the options that are significant to it if they are not critical. 822 4.1.4. Unidirectional Tunnels 824 Generally speaking, a Geneve tunnel is a unidirectional concept. IP 825 is not a connection oriented protocol and it is possible for two 826 endpoints to communicate with each other using different paths or to 827 have one side not transmit anything at all. As Geneve is an IP-based 828 protocol, the tunnel layer inherits these same characteristics. 830 It is possible for a tunnel to encapsulate a protocol, such as TCP, 831 which is connection oriented and maintains session state at that 832 layer. In addition, implementations MAY model Geneve tunnels as 833 connected, bidirectional links, such as to provide the abstraction of 834 a virtual port. In both of these cases, bidirectionality of the 835 tunnel is handled at a higher layer and does not affect the operation 836 of Geneve itself. 838 4.2. Constraints on Protocol Features 840 Geneve is intended to be flexible to a wide range of current and 841 future applications. As a result, certain constraints may be placed 842 on the use of metadata or other aspects of the protocol in order to 843 optimize for a particular use case. For example, some applications 844 may limit the types of options which are supported or enforce a 845 maximum number or length of options. Other applications may only 846 handle certain encapsulated payload types, such as Ethernet or IP. 847 This could be either globally throughout the system or, for example, 848 restricted to certain classes of devices or network paths. 850 These constraints may be communicated to tunnel endpoints either 851 explicitly through a control plane or implicitly by the nature of the 852 application. As Geneve is defined as a data plane protocol that is 853 control plane agnostic, the exact mechanism is not defined in this 854 document. 856 4.3. NIC Offloads 858 Modern NICs currently provide a variety of offloads to enable the 859 efficient processing of packets. The implementation of many of these 860 offloads requires only that the encapsulated packet be easily parsed 861 (for example, checksum offload). However, optimizations such as LSO 862 and LRO involve some processing of the options themselves since they 863 must be replicated/merged across multiple packets. In these 864 situations, it is desirable to not require changes to the offload 865 logic to handle the introduction of new options. To enable this, 866 some constraints are placed on the definitions of options to allow 867 for simple processing rules: 869 o When performing LSO, a NIC MUST replicate the entire Geneve header 870 and all options, including those unknown to the device, onto each 871 resulting segment. However, a given option definition may 872 override this rule and specify different behavior in supporting 873 devices. Conversely, when performing LRO, a NIC MAY assume that a 874 binary comparison of the options (including unknown options) is 875 sufficient to ensure equality and MAY merge packets with equal 876 Geneve headers. 878 o Options MUST NOT be reordered during the course of offload 879 processing, including when merging packets for the purpose of LRO. 881 o NICs performing offloads MUST NOT drop packets with unknown 882 options, including those marked as critical. 884 There is no requirement that a given implementation of Geneve employ 885 the offloads listed as examples above. However, as these offloads 886 are currently widely deployed in commercially available NICs, the 887 rules described here are intended to enable efficient handling of 888 current and future options across a variety of devices. 890 4.4. Inner VLAN Handling 892 Geneve is capable of encapsulating a wide range of protocols and 893 therefore a given implementation is likely to support only a small 894 subset of the possibilities. However, as Ethernet is expected to be 895 widely deployed, it is useful to describe the behavior of VLANs 896 inside encapsulated Ethernet frames. 898 As with any protocol, support for inner VLAN headers is OPTIONAL. In 899 many cases, the use of encapsulated VLANs may be disallowed due to 900 security or implementation considerations. However, in other cases 901 trunking of VLAN frames across a Geneve tunnel can prove useful. As 902 a result, the processing of inner VLAN tags upon ingress or egress 903 from a tunnel endpoint is based upon the configuration of the 904 endpoint and/or control plane and not explicitly defined as part of 905 the data format. 907 5. Interoperability Issues 909 Viewed exclusively from the data plane, Geneve does not introduce any 910 interoperability issues as it appears to most devices as UDP packets. 911 However, as there are already a number of tunnel protocols deployed 912 in network virtualization environments, there is a practical question 913 of transition and coexistence. 915 Since Geneve is a superset of the functionality of the three most 916 common protocols used for network virtualization (VXLAN, NVGRE, and 917 STT) it should be straightforward to port an existing control plane 918 to run on top of it with minimal effort. With both the old and new 919 packet formats supporting the same set of capabilities, there is no 920 need for a hard transition - endpoints directly communicating with 921 each other use any common protocol, which may be different even 922 within a single overall system. As transit devices are primarily 923 forwarding packets on the basis of the IP header, all protocols 924 appear similar and these devices do not introduce additional 925 interoperability concerns. 927 To assist with this transition, it is strongly suggested that 928 implementations support simultaneous operation of both Geneve and 929 existing tunnel protocols as it is expected to be common for a single 930 node to communicate with a mixture of other nodes. Eventually, older 931 protocols may be phased out as they are no longer in use. 933 6. Security Considerations 935 As UDP/IP packets, Geneve does not have any inherent security 936 mechanisms. As a result, an attacker with access to the underlay 937 network transporting the IP packets has the ability to snoop or 938 inject packets. Legitimate but malicious tunnel endpoints may also 939 spoof identifiers in the tunnel header to gain access to networks 940 owned by other tenants. 942 Within a particular security domain, such as a data center operated 943 by a single provider, the most common and highest performing security 944 mechanism is isolation of trusted components. Tunnel traffic can be 945 carried over a separate VLAN and filtered at any untrusted 946 boundaries. In addition, tunnel endpoints should only be operated in 947 environments controlled by the service provider, such as the 948 hypervisor itself rather than within a customer VM. 950 When crossing an untrusted link, such as the public Internet, IPsec 951 [RFC4301] may be used to provide authentication and/or encryption of 952 the IP packets formed as part of Geneve encapsulation. If the remote 953 tunnel endpoint is not completely trusted, for example it resides on 954 a customer premises, then it may also be necessary to sanitize any 955 tunnel metadata to prevent tenant-hopping attacks. 957 Geneve does not otherwise affect the security of the encapsulated 958 packets. 960 7. IANA Considerations 962 IANA has allocated UDP port 6081 as the well-known destination port 963 for Geneve. Upon publication, the registry should be updated to cite 964 this document. The original request was: 966 Service Name: geneve 967 Transport Protocol(s): UDP 968 Assignee: Jesse Gross 969 Contact: Jesse Gross 970 Description: Generic Network Virtualization Encapsulation (Geneve) 971 Reference: This document 972 Port Number: 6081 974 In addition, IANA is requested to create a "Geneve Option Class" 975 registry to allocate Option Classes. This shall be a registry of 976 16-bit hexadecimal values along with descriptive strings. The 977 identifiers 0x0-0xFF are to be reserved for standardized options for 978 allocation by IETF Review [RFC5226] and 0xFFF0-0xFFFF for 979 Experimental Use. Otherwise, identifiers are to be assigned to any 980 organization with an interest in creating Geneve options on a First 981 Come First Served basis. The registry is to be populated with the 982 following initial values: 984 +----------------+--------------------------------------+ 985 | Option Class | Description | 986 +----------------+--------------------------------------+ 987 | 0x0000..0x00FF | Unassigned - IETF Review | 988 | 0x0100 | Linux | 989 | 0x0101 | Open vSwitch | 990 | 0x0102 | Open Virtual Networking (OVN) | 991 | 0x0103 | In-band Network Telemetry (INT) | 992 | 0x0104 | VMware | 993 | 0x0105..0xFFEF | Unassigned - First Come First Served | 994 | 0xFFF0..FFFF | Experimental | 995 +----------------+--------------------------------------+ 997 8. Contributors 999 The following individuals were authors of an earlier version of this 1000 document and made significant contributions: 1002 Pankaj Garg 1003 Microsoft Corporation 1004 1 Microsoft Way 1005 Redmond, WA 98052 1006 USA 1008 Email: pankajg@microsoft.com 1010 Chris Wright 1011 Red Hat Inc. 1012 1801 Varsity Drive 1013 Raleigh, NC 27606 1014 USA 1016 Email: chrisw@redhat.com 1018 Puneet Agarwal 1019 Innovium, Inc. 1020 6001 America Center Drive 1021 San Jose, CA 95002 1022 USA 1024 Email: puneet@innovium.com 1025 Kenneth Duda 1026 Arista Networks 1027 5453 Great America Parkway 1028 Santa Clara, CA 95054 1029 USA 1031 Email: kduda@arista.com 1033 Dinesh G. Dutt 1034 Cumulus Networks 1035 140C S. Whisman Road 1036 Mountain View, CA 94041 1037 USA 1039 Email: ddutt@cumulusnetworks.com 1041 Jon Hudson 1042 Brocade Communications Systems, Inc. 1043 130 Holger Way 1044 San Jose, CA 95134 1045 USA 1047 Email: jon.hudson@gmail.com 1049 Ariel Hendel 1050 Broadcom Limited 1051 3151 Zanker Road 1052 San Jose, CA 95134 1053 USA 1055 Email: ariel.hendel@broadcom.com 1057 9. Acknowledgements 1059 The authors wish to thank Martin Casado, Bruce Davie and Dave Thaler 1060 for their input, feedback, and helpful suggestions. 1062 10. References 1064 10.1. Normative References 1066 [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, 1067 DOI 10.17487/RFC0768, August 1980, 1068 . 1070 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1071 Requirement Levels", BCP 14, RFC 2119, 1072 DOI 10.17487/RFC2119, March 1997, 1073 . 1075 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an 1076 IANA Considerations Section in RFCs", BCP 26, RFC 5226, 1077 DOI 10.17487/RFC5226, May 2008, 1078 . 1080 10.2. Informative References 1082 [ETYPES] The IEEE Registration Authority, "IEEE 802 Numbers", 2013, 1083 . 1086 [I-D.davie-stt] 1087 Davie, B. and J. Gross, "A Stateless Transport Tunneling 1088 Protocol for Network Virtualization (STT)", draft-davie- 1089 stt-08 (work in progress), April 2016. 1091 [I-D.ietf-nvo3-dataplane-requirements] 1092 Bitar, N., Lasserre, M., Balus, F., Morin, T., Jin, L., 1093 and B. Khasnabish, "NVO3 Data Plane Requirements", draft- 1094 ietf-nvo3-dataplane-requirements-03 (work in progress), 1095 April 2014. 1097 [IEEE.802.1Q-2014] 1098 IEEE, "IEEE Standard for Local and metropolitan area 1099 networks -- Bridges and Bridged Networks", IEEE 1100 Std 802.1Q, 2014. 1102 [RFC1191] Mogul, J. and S. Deering, "Path MTU discovery", RFC 1191, 1103 DOI 10.17487/RFC1191, November 1990, 1104 . 1106 [RFC1981] McCann, J., Deering, S., and J. Mogul, "Path MTU Discovery 1107 for IP version 6", RFC 1981, DOI 10.17487/RFC1981, August 1108 1996, . 1110 [RFC2983] Black, D., "Differentiated Services and Tunnels", 1111 RFC 2983, DOI 10.17487/RFC2983, October 2000, 1112 . 1114 [RFC3031] Rosen, E., Viswanathan, A., and R. Callon, "Multiprotocol 1115 Label Switching Architecture", RFC 3031, 1116 DOI 10.17487/RFC3031, January 2001, 1117 . 1119 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1120 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 1121 December 2005, . 1123 [RFC6040] Briscoe, B., "Tunnelling of Explicit Congestion 1124 Notification", RFC 6040, DOI 10.17487/RFC6040, November 1125 2010, . 1127 [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and 1128 UDP Checksums for Tunneled Packets", RFC 6935, 1129 DOI 10.17487/RFC6935, April 2013, 1130 . 1132 [RFC7348] Mahalingam, M., Dutt, D., Duda, K., Agarwal, P., Kreeger, 1133 L., Sridhar, T., Bursell, M., and C. Wright, "Virtual 1134 eXtensible Local Area Network (VXLAN): A Framework for 1135 Overlaying Virtualized Layer 2 Networks over Layer 3 1136 Networks", RFC 7348, DOI 10.17487/RFC7348, August 2014, 1137 . 1139 [RFC7365] Lasserre, M., Balus, F., Morin, T., Bitar, N., and Y. 1140 Rekhter, "Framework for Data Center (DC) Network 1141 Virtualization", RFC 7365, DOI 10.17487/RFC7365, October 1142 2014, . 1144 [RFC7637] Garg, P., Ed. and Y. Wang, Ed., "NVGRE: Network 1145 Virtualization Using Generic Routing Encapsulation", 1146 RFC 7637, DOI 10.17487/RFC7637, September 2015, 1147 . 1149 [VL2] Greenberg et al, , "VL2: A Scalable and Flexible Data 1150 Center Network", 2009. 1152 Proc. ACM SIGCOMM 2009 1154 Authors' Addresses 1156 Jesse Gross (editor) 1158 Email: jesse@kernel.org 1159 Ilango Ganga (editor) 1160 Intel Corporation 1161 2200 Mission College Blvd. 1162 Santa Clara, CA 95054 1163 USA 1165 Email: ilango.s.ganga@intel.com 1167 T. Sridhar (editor) 1168 VMware, Inc. 1169 3401 Hillview Ave. 1170 Palo Alto, CA 94304 1171 USA 1173 Email: tsridhar@vmware.com