Network Working Group E. Crabbe Internet-Draft Intended status: Standard Track L. Yong Huawei USA X. Xu Huawei Technologies T. Herbert Google Expires: July 2016 January 26, 2016 GRE-in-UDP Encapsulation draft-ietf-tsvwg-gre-in-udp-encap-09 Abstract This document describes a method of encapsulating network protocol packets within GRE and UDP headers. In this encapsulation, the source UDP port can be used as an entropy field for purposes of load balancing, while the protocol of the encapsulated packet in the GRE payload is identified by the GRE Protocol Type. This document specifies requirements for two applicability scenarios for this encapsulation: (1) General Internet and (2) well-managed operator networks; less restrictive requirements apply to the latter scenario by comparison to the former. Status of This Document This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://datatracker.ietf.org/drafts/current/. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." This Internet-Draft will expire on July 26,2016. Copyright Notice Copyright (c) 2015 IETF Trust and the persons identified as the document authors. All rights reserved. Crabbe, Yong, Xu, Herbert [Page 1] Internet-Draft GRE-in-UDP Encapsulation January 2016 This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1. Introduction...................................................3 1.1. Applicability Statement...................................4 1.2. GRE-in-UDP Tunnel Usage Requirements......................4 1.2.1. Requirements for Default GRE-in-UDP Tunnel...........5 1.2.2. Requirements Changes for WMON GRE-in-UDP Tunnel......5 2. Terminology....................................................6 2.1. Requirements Language.....................................6 3. Encapsulation in UDP...........................................6 3.1. IP Header.................................................9 3.2. UDP Header................................................9 3.2.1. Source Port..........................................9 3.2.2. Destination Port.....................................9 3.2.3. Checksum............................................10 3.2.4. Length..............................................10 3.3. GRE Header...............................................10 4. Encapsulation Process Procedures..............................11 4.1. MTU and Fragmentation....................................11 4.2. Middlebox Considerations.................................12 4.3. Differentiated Services and ECN Marking..................12 5. UDP Checksum Handling.........................................13 5.1. UDP Checksum with IPv4...................................13 5.2. UDP Checksum with IPv6...................................13 5.2.1. Middlebox Considerations............................16 6. Congestion Considerations.....................................17 7. Backward Compatibility........................................18 8. IANA Considerations...........................................19 9. Security Considerations.......................................19 10. Acknowledgements.............................................20 11. Contributors.................................................21 12. References...................................................22 12.1. Normative References....................................22 12.2. Informative References..................................23 13. Authors' Addresses...........................................24 Crabbe, Yong, Xu, Herbert [Page 2] Internet-Draft GRE-in-UDP Encapsulation January 2016 1. Introduction Load balancing, or more specifically statistical multiplexing of traffic using Equal Cost Multi-Path (ECMP) and/or Link Aggregation Groups (LAGs) in IP networks is a widely used technique for creating higher capacity networks out of lower capacity links. Most existing routers in IP networks are already capable of distributing IP traffic flows over ECMP paths and/or LAGs on the basis of a hash function performed on flow invariant fields in IP packet headers and their payload protocol headers. Specifically, when the IP payload is a User Datagram Protocol (UDP)[RFC768] or Transmission Control Protocol (TCP) [RFC793] packet, router hash functions frequently operate on the five-tuple of source IP address, destination IP address, source port, destination port, and protocol/next-header GRE encapsulation has been widely used for many applications. For example, to redirect IP traffic to traverse a different path instead of the default path in an operator network, to tunnel private network traffic over a public network by use of public IP network addresses, to tunnel IPv6 traffic over an IPv4 network, tunnel Ethernet traffic over IP networks [RFC7637], etc. Unfortunately, use of common GRE endpoints may reduce the entropy available for use in load balancing, especially in environments where the GRE Key field [RFC2890] is not readily available for use as entropy in forwarding decisions. This document defines a generic GRE-in-UDP encapsulation for tunneling network protocol packets across an IP network. The GRE header provides payload protocol type as an EtherType in the protocol type field [RFC2784][GREIPV6], and the UDP header provides additional entropy by way of its source port. GRE-in-UDP offers the additional possibility of using GRE across networks that might otherwise disallow it; for instance GRE-in-UDP may be used to bridge two islands where GRE is not used natively across the Internet. This encapsulation method requires no changes to the transit IP network. Hash functions in most existing IP routers may utilize and benefit from the use of a GRE-in-UDP tunnel without needing any change or upgrade to their ECMP implementation. The encapsulation mechanism is applicable to a variety of IP networks including Data Center and wide area networks. GRE-in-UDP encapsulation may be used to encapsulate already tunneled traffic, i.e. tunnel-in-tunnel. The tunneled traffic may use GRE-in- UDP or other tunnel encapsulation. In this case, GRE-in-UDP tunnel endpoints treat other tunnel endpoints as of the end hosts for the Crabbe, Yong, Xu, Herbert [Page 3] Internet-Draft GRE-in-UDP Encapsulation January 2016 traffic and do not differentiate such end hosts from other end hosts. 1.1. Applicability Statement GRE-in-UDP encapsulation applies to IPv4 and IPv6 networks including the Internet. When using GRE-in-UDP encapsulation, encapsulated traffic will be treated as a UDP application in an IP network. As such, GRE-in-UDP tunnel needs to meet UDP requirements specified in [RFC5405bis], which imposes limits on GRE-in-UDP tunnel usage. These limits may depend on both the network and the nature of the encapsulated traffic. For example, the GRE-in-UDP tunnel protocol does not provide any congestion control functionality beyond that of the encapsulated traffic. Therefore, GRE-in-UDP MUST be used only with congestion controlled traffic (e.g., IP traffic) and/or within a network that has the congestion management. [RFC5405bis] considers two types of applicability where IETF applications utilize UDP: 1) General Internet and 2) Controlled Environment. The controlled environment means within a single administrative domain or bilaterally agreed connection between domains. A network under controlled environment can be managed/operated to meet certain condition(s), which the general Internet can't. Tunnel protocol requirements under controlled environment can be less restrictive than the requirements in the general Internet. This document specifies GRE-in-UDP tunnel usage in the general Internet and GRE-in-UDP tunnel usage in the well-managed operator network that is an example of controlled environment. For the purpose of this document, a well-managed operator network is defined as an IP network that is traffic-engineered and/or otherwise managed (e.g., via use of traffic rate limiters) to avoid congestion. This document refers to the GRE-in-UDP tunnel usage in the general Internet as Default GRE-in-UDP Tunnel; the GRE-in-UDP tunnel usage in a well-managed operator network as WMON GRE-in-UDP Tunnel. 1.2. GRE-in-UDP Tunnel Usage Requirements The section summarizes GRE-in-UDP tunnel requirements. The requirements for Default GRE-in-UDP tunnel are listed in Section 1.2.1, which applies to a GRE-in-UDP tunnel over the general Internet; the relaxed requirements for WMON GRE-in-UDP Tunnel are listed in Section 1.2.2, which applies to a GRE-in-UDP tunnel within a well-managed operator network. These networks can be IPv4 or IPv6. Crabbe, Yong, Xu, Herbert [Page 4] Internet-Draft GRE-in-UDP Encapsulation January 2016 1.2.1. Requirements for Default GRE-in-UDP Tunnel The following is a summary of the GRE-in-UDP requirements for use over the general Internet: 1. UDP checksum SHOULD be used when encapsulating in IPv4. 2. UDP checksum MUST be used when encapsulating in IPv6. 3. GRE-in-UDP tunnel MUST NOT be used for traffic that has no congestion control. IP-traffic can be assumed to be congestion- controlled. GRE-in-UDP tunnels are not appropriate for other traffic that does not use congestion control. 4. UDP source port that is used for flow entropy SHOULD be set to a UDP ephemeral port (49152-65535). 5. UDP source port usage MUST be configurable so that a single value is used for all traffic in the tunnel (this disables use of the UDP source port to provide flow entropy). 6. For IPv6 delivery networks, the flow entropy SHOULD also be placed in the flow label field. 7. At the tunnel ingress, any fragmentation of the incoming packet (e.g., because the tunnel has an MTU that is smaller than the packet SHOULD be performed before encapsulation [RFC7588]. In addition, the tunnel ingress MUST apply the UDP checksum to all encapsulated fragments so that the tunnel egress can validate reassembly of the fragments, and SHOULD use the same source UDP port for all packet fragments to ensure the packet fragments traversing on the same path. 1.2.2. Requirements Changes for WMON GRE-in-UDP Tunnel The following lists the changed requirements for WMON GRE-in-UDP Tunnel that is used in a well-managed operator network; they replace requirements 1-3 listed in section 1.2.1. The requirements 4-8 in that section are unchanged for WMON GRE-in-UDP Tunnel. 1. UDP checksum MAY be used when encapsulating in IPv4. 2. Use of UDP checksum MUST be the default when encapsulating in IPv6. This default MAY be overridden via configuration of UDP zero- checksum mode. All usage of UDP zero-checksum mode with IPv6 is subject to the additional requirements specified in Section 5.2. Crabbe, Yong, Xu, Herbert [Page 5] Internet-Draft GRE-in-UDP Encapsulation January 2016 3. GRE-in-UDP tunnel MAY encapsulate traffic that is not congestion controlled. 2. Terminology The terms defined in [RFC768][RFC2784] are used in this document. Default GRE-in-UDP Tunnel: A GRE-in-UDP tunnel that can apply to the general Internet. WMON GRE-in-UDP Tunnel: A GRE-in-UDP tunnel that can only apply to a well-managed operator network that is defined in Section 1.1. 2.1. Requirements Language The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119]. 3. Encapsulation in UDP GRE-in-UDP encapsulation format is shown as follows: Crabbe, Yong, Xu, Herbert [Page 6] Internet-Draft GRE-in-UDP Encapsulation January 2016 0 1 2 3 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 IPv4 Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |Version| IHL |Type of Service| Total Length | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Identification |Flags| Fragment Offset | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Time to Live |Protcol=17(UDP)| Header Checksum | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Source IPv4 Address | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Destination IPv4 Address | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ UDP Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Source Port = XXXX | Dest Port = TBD | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | UDP Length | UDP Checksum | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ GRE Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |C| |K|S| Reserved0 | Ver | Protocol Type | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Checksum (optional) | Reserved1 (Optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Key (optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Sequence Number (optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 1 UDP+GRE Headers in IPv4 Crabbe, Yong, Xu, Herbert [Page 7] Internet-Draft GRE-in-UDP Encapsulation January 2016 0 1 2 3 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 IPv6 Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |Version| Traffic Class | Flow Label | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Payload Length | NxtHdr=17(UDP)| Hop Limit | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | + + | | + Outer Source IPv6 Address + | | + + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | + + | | + Outer Destination IPv6 Address + | | + + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ UDP Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Source Port = XXXX | Dest Port = TBD | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | UDP Length | UDP Checksum | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ GRE Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |C| |K|S| Reserved0 | Ver | Protocol Type | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Checksum (optional) | Reserved1 (Optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Key (optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Sequence Number (optional) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Figure 2 UDP+GRE Headers in IPv6 Crabbe, Yong, Xu, Herbert [Page 8] Internet-Draft GRE-in-UDP Encapsulation January 2016 The contents of the IP, UDP, and GRE headers that are relevant in this encapsulation are described below. 3.1. IP Header An encapsulator MUST encode its own IP address as the source IP address and the decapsulator's IP address as the destination IP address. The TTL field in the IP header MUST be set to a value appropriate for delivery of the encapsulated packet to the peer of the encapsulation. 3.2. UDP Header 3.2.1. Source Port The UDP source port contains a 16-bit entropy value that is generated by the encapsulator to identify a flow for the encapsulated packet. The port value SHOULD be within the ephemeral port range, i.e., 49152 to 65535, where the high order two bits of the port are set to one. This provides fourteen bits of entropy for the inner flow identifier. In the case that an encapsulator is unable to derive flow entropy from the payload header or the entropy usage has to be disabled for a purpose (see section 4.2), it SHOULD set a randomly selected constant value for UDP source port to avoid payload packet flow reordering, e.g., use of the system time to yield a value that is the range of entropy values. The source port value for a flow set by an encapsulator MAY change over the lifetime of the encapsulated flow. For instance, an encapsulator may change the assignment for Denial of Service (DOS) mitigation or as a means to effect routing through the ECMP network. An encapsulator SHOULD NOT change the source port selected for a flow more than once every thirty seconds. For IPv6 delivery network, if IPv6 flow label load balancing is supported [RFC6438], the flow entropy SHOULD also be placed in the flow label field. How an encapsulator generates flow entropy from the payload is outside the scope of this document. 3.2.2. Destination Port The destination port of the UDP header is set the GRE-in-UDP port or GRE-UDP-DTLS (TBD) (see Section 8). Crabbe, Yong, Xu, Herbert [Page 9] Internet-Draft GRE-in-UDP Encapsulation January 2016 3.2.3. Checksum The UDP checksum is set and processed per [RFC768] and [RFC1122] for IPv4, and [RFC2460] for IPv6. Requirements for checksum handling and use of zero UDP checksums are detailed in Section 5. 3.2.4. Length The usage of this field is in accordance with the current UDP specification in [RFC768]. This length will include the UDP header (eight bytes), GRE header, and the GRE payload (encapsulated packet). 3.3. GRE Header An encapsulator sets the protocol type (EtherType) of the packet being encapsulated in the GRE Protocol Type field. An encapsulator may set the GRE Key Present, Sequence Number Present, and Checksum Present bits and associated fields in the GRE header as defined by [RFC2784] and [RFC2890]. The reserved bits, i.e., Reserved0, SHOULD be set zero. The GRE checksum MAY be enabled to protect the GRE header and payload. An encapsulator SHOULD NOT enable both the GRE checksum and UDP checksum simultaneously as this would be mostly redundant. Since the UDP checksum covers more of the packet including the GRE header and payload, the UDP checksum SHOULD have preference to using GRE checksum. The GRE checksum SHOULD be used for the payload integrity check when use of UDP zero-checksum. An implementation MAY use the GRE keyid to authenticate the encapsulator. (See Security Section) In this model, a shared value is either configured or negotiated between an encapsulator and decapsulator. When a decapsulator determines a presented keyid is not valid for the source, the packet MUST be dropped. Although GRE-in-UDP encapsulation protocol uses both UDP header and GRE header, it is one tunnel encapsulation protocol. GRE and UDP headers MUST be applied and removed as a pair at the encapsulation and decapsulation points. This specification does not support UDP encapsulation of a GRE header where that GRE header is applied or removed at a network node other than the UDP tunnel ingress or egress. Crabbe, Yong, Xu, Herbert [Page 10] Internet-Draft GRE-in-UDP Encapsulation January 2016 4. Encapsulation Process Procedures The procedures specified in this section apply to both Default GRE- in-UDP tunnel and WMON GRE-in-UDP tunnel. The GRE-in-UDP encapsulation allows encapsulated packets to be forwarded through "GRE-in-UDP tunnels". When performing GRE-in-UDP encapsulation by the encapsulator, the entropy value is generated by the encapsulator and then be filled in the Source Port field of the UDP header. The Destination Port field is set to a value (TBD) to indicate that the UDP tunnel payload is a GRE packet. The Protocol Type header field in GRE header is set to the EtherType value corresponding to the protocol of the encapsulated packet. Intermediate routers, upon receiving these UDP encapsulated packets, could balance these packets based on the hash of the five-tuple of UDP packets. Upon receiving these UDP encapsulated packets, the decapsulator decapsulates them by removing the UDP and GRE headers and then processes them accordingly. GRE-in-UDP allows encapsulation of unicast, broadcast, or multicast traffic. Entropy may be generated from the header of encapsulated unicast or broadcast/multicast packets at an encapsulator. The mapping mechanism between the encapsulated multicast traffic and the multicast capability in the IP network is transparent and independent to the encapsulation and is otherwise outside the scope of this document. To provide entropy for ECMP, GRE-in-UDP does not rely on GRE keep- alive. It is RECOMMENED no use of GRE keep-alive in the GRE-in-UDP tunnel. This aligns with middlebox traversal guidelines in Section 3.5 of [RFC5405bis]. 4.1. MTU and Fragmentation Regarding packet fragmentation, an encapsulator/decapsulator SHOULD be compliant with [RFC7588] and the encapsulator MUST fragment the packet before the encapsulation. For this case, the MTU for the payload SHOULD be less or equal to the PMTU associated with the path between the GRE ingress and the GRE egress nodes minus the GRE and UDP overhead, assuming the egress resemble MTU is larger than PMTU. When applying payload fragment, the UDP checksum MUST be used so that the receiving endpoint can validate reassembly of the fragments; the same src UDP port SHOULD be used for all packet fragments to Crabbe, Yong, Xu, Herbert [Page 11] Internet-Draft GRE-in-UDP Encapsulation January 2016 ensure the transit routers will forward the fragments on the same path. If a tunnel operator is able to control the payload MTU size, the tunnel operator SHOULD factor in the additional bytes of tunnel overhead when considering the MTU size to avoid the likelihood of fragmentation. 4.2. Middlebox Considerations The Source Port number of the UDP header is pertinent to the middlebox behavior. Network Address/Port Translator (NAPT) is the most commonly deployed Network Address Translation (NAT) device [RFC4787]. An NAPT device establishes a NAT session to translate the {private IP address, private source port number} tuple to a {public IP address, public source port number} tuple, and vice versa, for the duration of the UDP session. This provides a UDP application with the "NAT-pass-through" function. NAPT allows multiple internal hosts to share a single public IP address. The port number, i.e., the UDP Source Port number, is used as the demultiplexer of the multiple internal hosts. However, the above NAPT behaviors conflict with the behavior that the UDP source port number is used as entropy in GRE-in-UDP tunnel. Each UDP tunnel is unidirectional, as GRE-in-UDP traffic is sent to the GRE-in-UDP Destination Port (TBD), and in particular, is never sent back to any port used as a UDP Source Port (which serves solely as a source of entropy). It is common that a middlebox (e.g., firewall) assume that bidirectional traffic uses a common pair of UDP ports. This assumption also conflicts with the use of the UDP source port number as entropy. Hence, use UDP src port for entropy may impact middlebox behavior. If a GRE-in-UDP tunnel is expected to pass a middlebox, to avoid the impact, the operator either disable UDP source port for entropy or configure the middlebox to deal with the UDP source port variation. 4.3. Differentiated Services and ECN Marking To ensure that tunneled traffic gets the same treatment over the IP network, prior to the encapsulation process, an encapsulator should process the payload to get the proper parameters to fill into the IP header such as DiffServ [RFC2983]. Encapsulation end points that support Explicit Congestion Notification (ECN) must use the method described in [RFC6040] for ECN marking propagation. The congestion control process is outside of the scope of this document. Crabbe, Yong, Xu, Herbert [Page 12] Internet-Draft GRE-in-UDP Encapsulation January 2016 5. UDP Checksum Handling 5.1. UDP Checksum with IPv4 Default GRE-in-UDP Tunnel SHOULD perform UDP checksum. WMON GRE-in- UDP Tunnel MAY perform UDP checksum. For UDP in IPv4, the UDP checksum MUST be processed as specified in [RFC768] and [RFC1122] for both transmit and receive. The IPv4 header includes a checksum which protects against mis-delivery of the packet due to corruption of IP addresses. The UDP checksum potentially provides protection against corruption of the UDP header, GRE header, and GRE payload. Enabling or disabling the use of checksums is a deployment consideration that should take into account the risk and effects of packet corruption, and whether the packets in the network are protected by other, possibly stronger mechanisms such as the Ethernet CRC. When a decapsulator receives a packet, the UDP checksum field MUST be processed. If the UDP checksum is non-zero, the decapsulator MUST verify the checksum before accepting the packet. By default a decapsulator SHOULD accept UDP packets with a zero checksum. A node MAY be configured to disallow zero checksums per [RFC1122]; this may be done selectively, for instance disallowing zero checksums from certain hosts that are known to be sending over paths subject to packet corruption. If verification of a non-zero checksum fails, a decapsulator lacks the capability to verify a non-zero checksum, or a packet with a zero-checksum was received and the decapsulator is configured to disallow, the packet MUST be dropped and an event MAY be logged. 5.2. UDP Checksum with IPv6 For UDP in IPv6, the UDP checksum MUST be processed as specified in [RFC768] and [RFC2460] for both transmit and receive. When UDP is used over IPv6, the UDP checksum is relied upon to protect both the IPv6 and UDP headers from corruption. As such, Default GRE-in-UDP Tunnel MUST perform UDP checksum; WMON GRE-in-UDP Tunnel MAY be configured with the UDP zero-checksum mode if the well-managed operator network meets at least one of following conditions and/or within a set of closely cooperating well-managed operator network administrations (such as network operators who have agreed to work together in order to jointly provide specific services). Crabbe, Yong, Xu, Herbert [Page 13] Internet-Draft GRE-in-UDP Encapsulation January 2016 a. Under single administrative control where it is known (perhaps through knowledge of equipment types and lower layer checks) that packet corruption is exceptionally unlikely and where the operator is willing to take the risk of undetected packet corruption. b. Under single administrative control (such as within a single operator's network) where it is judged through observational measurements (perhaps of historic or current traffic flows that use a non-zero checksum) that the level of packet corruption is tolerably low and where the operator is willing to take the risk of undetected packet corruption. c. Carrying applications that are tolerant of mis-delivered or corrupted packets (perhaps through higher layer checksum, validation, and retransmission or transmission redundancy) where the operator is willing to rely on the applications using the tunnel to survive any corrupt packets. As such, Default GRE-in-UDP tunnel MUST perform UDP checksum; WMON GRE-in-UDP Tunnel MUST perform UDP checksum by default, this default MAY be overridden via configuration of UDP zero-checksum mode subject to the additional requirements specified below. The following additional requirements apply to WMON GRE-in-UDP Tunnel that use UDP zero-checksum mode: a. Use of the UDP checksum with IPv6 MUST be the default configuration of all GRE-in-UDP tunnels. b. The GRE-in-UDP tunnel implementation MUST comply with all requirements specified in Section 4 of [RFC6936] and with requirement 1 specified in Section 5 of [RFC6936]. c. The tunnel decapsulator SHOULD only allow the use of UDP zero- checksum mode for IPv6 on a single received UDP Destination Port regardless of the encapsulator. The motivation for this requirement is possible corruption of the UDP Destination Port, which may cause packet delivery to the wrong UDP port. If that other UDP port requires the UDP checksum, the mis-delivered packet will be discarded. Crabbe, Yong, Xu, Herbert [Page 14] Internet-Draft GRE-in-UDP Encapsulation January 2016 d. It is RECOMMENDED that UDP zero-checksum selectively be enabled for certain source addresses. The tunnel decapsulator MUST check that the source and destination IPv6 addresses are valid for the GRE-in-UDP tunnel on which the packet was received if that tunnel uses UDP zero-checksum mode and discard any packet for which this check fails. e. The tunnel encapsulator SHOULD use different IPv6 addresses for each GRE-in-UDP tunnel that uses UDP zero-checksum mode regardless of the decapsulator in order to strengthen the decapsulator's check of the IPv6 source address (i.e., the same IPv6 source address SHOULD NOT be used with more than one IPv6 destination address, independent of whether that destination address is a unicast or multicast address). When this is not possible, it is RECOMMENDED to use each source IPv6 address for as few UDP zero-checksum mode GRE-in-UDP tunnels as is feasible. Note that if UDP checksum is used, such restriction is not necessary. f. When any middlebox exists on the path of GRE-in-UDP tunnel, it is RECOMMENDED to use the default mode, i.e. use UDP checksum, to reduce the chance that the encapsulated packets to be dropped. g. Any middlebox for UDP zero-checksum mode for IPv6 MUST comply with requirement 1 and 8-10 in Section 5 of [RFC6936]. h. Measures SHOULD be taken to prevent IPv6 traffic with zero UDP checksums from "escaping" to the general Internet; see Section 6 for examples of such measures. i. IPv6 traffic with zero UDP checksums MUST be actively monitored for errors by the network operator. For example, the operator may monitor Ethernet layer packet error rates. j. If a packet with a non-zero checksum is received, the checksum MUST be verified before accepting the packet. This is regardless of whether the tunnel encapsulator and decapsulator have been configured with UDP zero-checksum mode. The above requirements do not change either the requirements specified in [RFC2460] as modified by [RFC6935] or the requirements specified in [RFC6936]. The requirement to check the source IPv6 address in addition to the destination IPv6 address, plus the strong recommendation against reuse of source IPv6 addresses among GRE-in-UDP tunnels collectively Crabbe, Yong, Xu, Herbert [Page 15] Internet-Draft GRE-in-UDP Encapsulation January 2016 provide some mitigation for the absence of UDP checksum coverage of the IPv6 header. Additional assurance is provided by the well- managed operator network that satisfies at least one of three conditions listed in this section (beginning). Hence GRE-in-UDP is suitable for transmission over lower layers in the well-managed operator networks that are allowed by the exceptions stated above and the rate of corruption of the inner IP packet on such networks is not expected to increase by comparison to GRE traffic that is not encapsulated in UDP. For these reasons, GRE-in-UDP does not provide an additional integrity check except when GRE checksum is used when UDP zero-checksum mode is used with IPv6, and this design is in accordance with requirements 2, 3 and 5 specified in Section 5 of [RFC6936]. GRE does not accumulate incorrect state as a consequence of GRE header corruption. A corrupt GRE packet may result in either packet discard or forwarding of the packet without accumulation of GRE state. Active monitoring of GRE-in-UDP traffic for errors is REQUIRED as occurrence of errors will result in some accumulation of error information outside the protocol for operational and management purposes. This design is in accordance with requirement 4 specified in Section 5 of [RFC6936]. The remaining requirements specified in Section 5 of [RFC6936] are inapplicable to GRE-in-UDP. Requirements 6 and 7 do not apply because GRE does not have a GRE-generic control feedback mechanism. Requirements 8-10 are middlebox requirements that do not apply to GRE-in-UDP tunnel endpoints, but see Section 5.2.1 for further middle box discussion. It is worth mentioning that the use of a zero UDP checksum should present the equivalent risk of undetected packet corruption when sending similar packet using GRE-in-IPv6 without UDP [GREIPV6] and without GRE checksums. In summary, WMON GRE-in-UDP Tunnel is allowed to use UDP-zero- checksum mode for IPv6, when additional requirements stated above are provided. Otherwise the UDP checksum MUST be used for IPv6 as specified in [RFC768] and [RFC2460]. Use of GRE checksum favors when the UDP checksum is not used. 5.2.1. Middlebox Considerations IPv6 datagrams with a zero UDP checksum will not be passed by any middlebox that validates the checksum based on [RFC2460] or that updates the UDP checksum field, such as NATs or firewalls. Changing Crabbe, Yong, Xu, Herbert [Page 16] Internet-Draft GRE-in-UDP Encapsulation January 2016 this behavior would require such middleboxes to be updated to correctly handle datagrams with zero UDP checksums. The GRE-in-UDP encapsulation does not provide a mechanism to safely fall back to using a checksum when a path change occurs redirecting a tunnel over a path that includes a middlebox that discards IPv6 datagrams with a zero UDP checksum. In this case the GRE-in-UDP tunnel will be black- holed by that middlebox. As such, when any middlebox exists on the path of GRE-in-UDP tunnel, it is RECOMMENDED to use the UDP checksum to reduce the chance that the encapsulated packets to be dropped. Recommended changes to allow firewalls, NATs and other middleboxes to support use of an IPv6 zero UDP checksum are described in Section 5 of [RFC6936]. 6. Congestion Considerations Section 3.1.3 of [RFC5405] discussed the congestion implications of UDP tunnels. As discussed in [RFC5405], because other flows can share the path with one or more UDP tunnels, congestion control [RFC2914] needs to be considered. The impact of congestion must be considered both in terms of the effect on the rest of the network of a UDP tunnel that is consuming excessive capacity, and in terms of the effect on the flows using the UDP tunnels. The potential impact of congestion from a UDP tunnel depends upon what sort of traffic is carried over the tunnel, as well as the path of the tunnel. In many cases, GRE-in-UDP is used to carry IP traffic. IP traffic is generally assumed to be congestion controlled, and thus a tunnel carrying general IP traffic generally does not need additional congestion control mechanisms. However, GRE-in-UDP tunnel can be used in some cases to carry traffic that is not necessarily congestion controlled. For example, GRE-in-UDP may be used to carry MPLS that carries pseudowire or VPN traffic where specific bandwidth guarantees are provided to each pseudowire or to each VPN. In such cases, network operators may avoid congestion by careful provisioning of their networks, by rate limiting of user data traffic, and traffic engineer according to path capacity. For this reason, GRE-in-UDP tunnel MUST be used within a single operator's network that utilizes careful provisioning (e.g., rate limiting at the entries of the network while over-provisioning network capacity) to ensure against congestion, or within a limited number of networks whose operators closely cooperate in order to jointly provide this same careful provisioning. Crabbe, Yong, Xu, Herbert [Page 17] Internet-Draft GRE-in-UDP Encapsulation January 2016 The default GRE-in-UDP tunnel can be used to carry IP traffic that is known to be congestion controlled on the Internet. Internet IP traffic is generally assumed to be congestion-controlled. The default GRE-in-UDP tunnel MUST NOT be used over the general Internet, or over non-cooperating network operators, to carry traffic that is not congestion-controlled. WMON GRE-in-UDP Tunnel is used within a well-managed operator network so that it can carry the traffic that is not necessary congestion controlled. Measures SHOULD be taken to prevent non- congestion-controlled GRE-in-UDP traffic from "escaping" to the general Internet, e.g.: o Physical or logical isolation of the links carrying GRE-in-UDP from the general Internet. o Deployment of packet filters that block the UDP ports assigned for GRE-in-UDP. o Imposition of restrictions on GRE-in-UDP traffic by software tools used to set up GRE-in-UDP tunnels between specific end systems (as might be used within a single data center). For examples, a GRE-in-UDP tunnel only carries IP traffic or a GRE- in-UDP tunnel supports NVEGRE encapsulation only (Although the payload type is Ethernet in NVGRE, NVGRE protocol mandates that the payload of Ethernet is IP). o Use of a "Circuit Breaker" for the tunneled traffic as described in [CB]. 7. Backward Compatibility In general, tunnel ingress routers have to be upgraded in order to support the encapsulations described in this document. No change is required at transit routers to support forwarding of the encapsulation described in this document. If a router that is intended for use as a decapsulator does not support or enable GRE-in-UDP encapsulation described in this document, it will not be listening on the destination port (TBD). In these cases, the router will conform to normal UDP processing and respond to an encapsulator with an ICMP message indicating "port unreachable" according to [RFC792]. Upon receiving this ICMP message, the node MUST NOT continue to use GRE-in-UDP encapsulation toward this peer without management intervention. Crabbe, Yong, Xu, Herbert [Page 18] Internet-Draft GRE-in-UDP Encapsulation January 2016 8. IANA Considerations IANA is requested to make the following allocations: One UDP destination port number for the indication of GRE Service Name: GRE-in-UDP Transport Protocol(s): UDP Assignee: IESG Contact: IETF Chair Description: GRE-in-UDP Encapsulation Reference: [This.I-D] Port Number: TBD Service Code: N/A Known Unauthorized Uses: N/A Assignment Notes: N/A One UDP destination port number for the indication of GRE with DTLS Service Name: GRE-UDP-DTLS Transport Protocol(s): UDP Assignee: IESG Contact: IETF Chair Description: GRE-in-UDP Encapsulation with DTLS Reference: [This.I-D] Port Number: TBD Service Code: N/A Known Unauthorized Uses: N/A Assignment Notes: N/A 9. Security Considerations GRE-in-UDP encapsulation does not affect security for the payload protocol. When using GRE-in-UDP, Network Security in a network is mostly equivalent to that of a network using GRE. Datagram Transport Layer Security (DTLS) [RFC6347] can be used for application security and can preserve network and transport layer protocol information. Specifically, if DTLS is used to secure the GRE-in-UDP tunnel, the destination port of the UDP header MUST be set to an IANA-assigned value (TBD2) indicating GRE-in-UDP with DTLS, and that UDP port MUST NOT be used for other traffic. The UDP Crabbe, Yong, Xu, Herbert [Page 19] Internet-Draft GRE-in-UDP Encapsulation January 2016 source port field can still be used to add entropy, e.g., for load- sharing purposes. DTLS usage is limited to a single DTLS session for any specific tunnel encapsulator/ decapsulator pair (identified by source and destination IP addresses). Both IP addresses MUST be unicast addresses - multicast traffic is not supported when DTLS is used. A GRE-in-UDP tunnel decapsulator that supports DTLS is expected to be able to establish DTLS sessions with multiple tunnel encapsulators, and likewise an GRE-in-UDP tunnel encapsulator is expected to be able to establish DTLS sessions with multiple decapsulators (although different source and/or destination IP addresses may be involved -see Section 5.2 for discussion of one situation where use of different source IP addresses is important). In the case that UDP source port for entropy usage is disabled, a random port SHOULD be selected in order to minimize the vulnerability to off-path attacks.[RFC6056] The random port may also be periodically changed to mitigate certain denial of service attacks as mentioned in Section 3.2. Using one standardized value in UDP destination port for an encapsulation indication may increase the vulnerability of off-path attack. To overcome this, an alternate port may be agreed upon to use between an encapsulator and decapsulator [RFC6056]. How the encapsulator end points communicate the value is outside scope of this document. This document does not require that decapsulator validates the IP source address of the tunneled packets (with the exception that the IPv6 source address MUST be validated when UDP zero-checksum mode is used with IPv6), but it should be understood that failure to do so presupposes that there is effective destination-based (or a combination of source-based and destination-based) filtering at the boundaries. Corruption of GRE header can cause a privacy and security concern for some applications that rely on the key field for traffic segregation. When GRE key field is used for privacy and security, ether UDP checksum or GRE checksum SHOULD be used for GRE-in-UDP with both IPv4 and IPv6, and in particular, when UDP zero-checksum mode is used, GRE checksum SHOULD be used. 10. Acknowledgements Authors like to thank Vivek Kumar, Ron Bonica, Joe Touch, Ruediger Geib, Lar Edds, Lloyd Wood, Bob Briscoe, and many others for their review and valuable input on this draft. Crabbe, Yong, Xu, Herbert [Page 20] Internet-Draft GRE-in-UDP Encapsulation January 2016 Thank the design team led by David Black (members: Ross Callon, Gorry Fairhurst, Xiaohu Xu, Lucy Yong) to efficiently work out the descriptions for the congestion considerations and IPv6 UDP zero checksum. 11. Contributors The following people all contributed significantly to this document and are listed below in alphabetical order: David Black EMC Corporation 176 South Street Hopkinton, MA 01748 USA Email: david.black@emc.com Ross Callon Juniper Networks 10 Technology Park Drive Westford, MA 01886 USA Email: rcallon@juniper.net John E. Drake Juniper Networks Email: jdrake@juniper.net Gorry Fairhurst University of Aberdeen Email: gorry@erg.abdn.ac.uk Yongbing Fan China Telecom Guangzhou, China. Phone: +86 20 38639121 Email: fanyb@gsta.com Crabbe, Yong, Xu, Herbert [Page 21] Internet-Draft GRE-in-UDP Encapsulation January 2016 Adrian Farrel Juniper Networks Email: adrian@olddog.co.uk Vishwas Manral Hewlett-Packard Corp. 3000 Hanover St, Palo Alto. Email: vishwas.manral@hp.com Carlos Pignataro Cisco Systems 7200-12 Kit Creek Road Research Triangle Park, NC 27709 USA EMail: cpignata@cisco.com 12. References 12.1. Normative References [RFC768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, August 1980. [RFC1122] Braden, R., "Requirements for Internet Hosts -- Communication Layers", RFC1122, October 1989. [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC2119, March 1997. [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P. Traina, "Generic Routing Encapsulation (GRE)", RFC 2784, March 2000. [RFC2890] Dommety, G., "Key and Sequence Number Extensions to GRE", RFC2890, September 2000. [RFC5405bis] Eggert, L., "Unicast UDP Usage Guideline for Application Designers", draft-ietf-tsvwg-rfc5405bis, work in progress. Crabbe, Yong, Xu, Herbert [Page 22] Internet-Draft GRE-in-UDP Encapsulation January 2016 [RFC6040] Briscoe, B., "Tunneling of Explicit Congestion Notification", RFC6040, November 2010. [RFC6347] Rescoria, E., Modadugu, N., "Datagram Transport Layer Security Version 1.2", RFC6347, 2012. [RFC6438] Carpenter, B., Amante, S., "Using the IPv6 Flow Label for Equal Cost Multipath Routing and Link Aggregation in tunnels", RFC6438, November, 2011. [RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and UDP Checksums for Tunneled Packets", RFC 6935, April 2013. [RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement for the Use of IPv6 UDP Datagrams with Zero Checksums", RFC 6936, April 2013. 12.2. Informative References [RFC792] Postel, J., "Internet Control Message Protocol", STD 5, RFC 792, September 1981. [RFC793] DARPA, "Transmission Control Protocol", RFC793, September 1981. [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 (IPv6) Specification", RFC 2460, December 1998. [RFC2914] Floyd, S.,"Congestion Control Principles", RFC2914, September 2000. [RFC2983] Black, D., "Differentiated Services and Tunnels", RFC2983, October 2000. [RFC3931] Lau, J., Townsley, M., and I. Goyret, "Layer Two Tunneling Protocol - Version 3 (L2TPv3)", RFC 3931, March 2005. [RFC4023] Worster, T., Rekhter, Y., and E. Rosen, "Encapsulating MPLS in IP or Generic Routing Encapsulation (GRE)", RFC 4023, March 2005. [RFC6056] Larsen, M. and Gont, F., "Recommendations for Transport- Protocol Port Randomization", RFC6056, January 2011. [RFC6438] Carpenter, B., Amante, S., "Using the Ipv6 Flow Label for Equal Cost Multipath Routing and Link Aggreation in Tunnels", RFC6438, November 2011. Crabbe, Yong, Xu, Herbert [Page 23] Internet-Draft GRE-in-UDP Encapsulation January 2016 [RFC7588] Bonica, R., "A Fragmentation Strategy for Generic Routing Encapsulation (GRE)", RFC7588, July 2015. [RFC7637] Garg, P. and Wang, Y., "NVGRE: Network Virtualization Using Generic Routing Encapsulation", RFC7637, September 2015. [CB] Fairhurst, G., "Network Transport Circuit Breakers", draft-ietf-tsvwg-circuit-breaker-04, work in progress. [GREIPV6] Pignataro, C., Bonica, R., Krishnan, S., "IPv6 Support for Generic Routing Encapsulation (GRE)", draft-ietf-intarea- gre-ipv6, work in progress. [TUNNEL] Touch, J. and Townsley, M., "IP Tunnels in the Internet Architecture", draft-ietf-intarea-tunnels-01.txt, work in progress. 13. Authors' Addresses Edward Crabbe Email: edward.crabbe@gmail.com Lucy Yong Huawei Technologies, USA Email: lucy.yong@huawei.com Xiaohu Xu Huawei Technologies, Beijing, China Email: xuxiaohu@huawei.com Tom Herbert Google 1600 Amphitheatre Parkway Mountain View, CA Email : tom@herbertland.com Crabbe, Yong, Xu, Herbert [Page 24]