Network working group L. Yong Internet Draft X. Xu Category: Standard Track Huawei Expires: November 2013 May 21, 2013 NVGRE and VXLAN Encapsulation for L3VPN Extension draft-yong-l3vpn-nvgre-vxlan-encap-01 Status of this Memo This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on November 21, 2013. Copyright Notice Copyright (c) 2009 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (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. Yong & Xu, et al. [Page 1] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 Abstract This document specifies NVGRE and VXLAN encapsulation for L3VPN Extension. Both NVGRE and VXLAN are originally designed for L2 overlay. The draft proposes the enhancement on both to allow L3 overlay completely decoupled from the L2 overlay in terms of encoding schema and data processing. Conventions used in this document The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC-2119 [RFC2119]. Table of Contents 1. Introduction...................................................3 2. NVGRE and VXLAN Encapsulation for L3VPN Extension..............3 2.1. NVGRE Enhancement for L3VPN Extension.....................3 2.2. VXLAN Enhancement for L3VPN Extension.....................4 2.3. BGP Tunnel Encapsulation Attribute Announcement...........5 2.4. Benefits..................................................5 3. Security Considerations........................................6 4. IANA Considerations............................................6 5. References.....................................................6 5.1. Normative References......................................6 5.2. Informative References....................................6 Yong & Xu [Page 2] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 1. Introduction DC Network Virtualization requires that a network virtual overlay may run at L2 or L3 [NVO3FRWK]. It is desirable to extend BGP/MPLS L3VPN [RFC4364] as one DC Network Virtualization solution. This implies that the PE component in the L3VPN architecture may be deployed in DC and reside on a same server. To complement current practical server implementations, the L3VPN extension [DRAO] [VAN] further proposes utilizing NVGRE [NVGRE] and/or VXLAN [VXLAN] data encapsulation formats for an L3VPN. This approach brings the advantage to use a unified solution for both L2 overlay [EVPN] and L3 overlay services. However, both NVGRE and VXLAN are originally designed as an L2 overlay data encapsulation in which the inner header MUST be an Ethernet header. This document proposes an enhancement to NVGRE and VXLAN encapsulation formats that allow the same data encapsulation semantics for both L2 overlay and L3 overlay services. The benefits of this usage include maintaining L3VPN natively and decoupling it from the L2 overlay completely. 2. NVGRE and VXLAN Encapsulation for L3VPN Extension A BGP/MPLS L3VPN [RFC4364] requires that the inner header is IP header, either IPv4 or IPv6. Both NVGRE [NVGRE] and VXLAN [VXLAN] encapsulation schema require that the inner header MUST be an Ethernet header. The solution of [DRAO] suggests several options to fill the MAC address fields when using NVGER and/or VXLAN in a BGP based L3VPN, although inner Ethernet address is not relevant to the L3VPN mechanism. These options make L3 overlay still coupled with L2 overlay one way or another [DRAO]. 2.1. NVGRE Enhancement for L3VPN Extension NVGER [NVGRE] leverages the GRE protocol [RFC2890] and specifies that the protocol type field in the GRE header MUST be filled with the value of 0x6558, which means for Transparent Ethernet. This draft proposes to allow the protocol type field to be either the value of 0x6558 or 0x0800. The value 0x0800 means IP payload [RFC3232]. The value 0x6558 MUST be used if the inner header is Ethernet header. The value 0x0800 MUST be used if the inner header is IP header. Furthermore, the version field in IP header indicates if the IP header is IPv4 or IPv6. When L3VPN Extension uses NVGRE encapsulation, it MUST use the value of 0x0800 in the protocol type Yong & Xu [Page 3] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 field and encode IP header as the inner header. Other fields in the outer header of the NVGRE remain the same. 2.2. VXLAN Enhancement for L3VPN Extension This document proposes adding a protocol type field in the VXLAN header as indicated below. It takes 16 bits from the reserved 24 bits as the protocol type field. For L2 overlay encapsulation, the protocol type field MUST be filled with the value of 0x6558. For L3 overlay encapsulation, the protocol type field MUST be filled with the value of 0x0800, which means IP header [RFC3232]; inner header MUST be IP header as shown below. The remained 8 reserved bit MUST be filled with zero. 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 Outer Ethernet Header: As described in VXLAN [VXLAN] Outer IP Header: As described in VXLAN [VXLAN] Outer UDP Header: As described in VXLAN [VXLAN] VXLAN Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |R|R|R|R|I|R|R|R| Reserved | Protocol Type = 0x6558/0x0800 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | VXLAN Network Identifier (VNI) | Reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Inner Header: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Ethernet header or IP Header ~ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ When L3VPN Extension uses VXLAN encapsulation, it MUST use the value of 0x0800 in the protocol type field and encode IP header as the inner header. Other fields in the outer header and the VXLAN header remain the same. For an L2 overlay, the protocol type value MUST be 0x6558, and inner header MUST be Ethernet header and use the same format as in [VXLAN]. For the backward compatibility, the value 0x0000 MUST be treated as Ethernet payload too. Other options to achieve this are either to specify another UDP port for L3 Overlay or to use one or more of reserved bits to indicate inner header type. For example, use the bit 6 in the VXLAN header. Yong & Xu [Page 4] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 However, UDP based option makes the overlay tied into underlay, i.e. rely on the underlay header to carry the overlay header type, and the bit option has limited space for future expansion and makes NVGRE and VXLAN use of different semantics to indicate inner header. Memo: to be consistent with ETHER Type, the Protocol Type MAY use different codes for IPv4 (0x800) and IPv6 (0x86dd) payload respectively. 2.3. BGP Tunnel Encapsulation Attribute Announcement RFC5512 [RFC5512] defines BGP tunnel encapsulation sub-TLVs and specifies a method for a BGP speaker to signal the tunnel encapsulation attributes to a peer. The BGP Remote-Next-Hop [VAN] uses the same sub-TLV to specify the remote next-hop tunnel encapsulation type. Both NVGRE and VXLAN are tunnel encapsulation options. This document suggests when specifying the tunnel encapsulation in the encapsulation sub-TLV, the egress router also specifies the proper protocol type in the protocol type sub-TLV [RFC5512]. The protocol type sub-TLV for IP or Ethernet overlay will be IPv4 (protocol type = 0x0800), IPv6 (0x86dd), and Ethernet (0x6558), respectively. The ingress router MUST encapsulate the packets with the encapsulation type that egress BGP speaker signals. 2.4. Benefits The enhancement on NVGRE and VXLAN encapsulation formats brings some beneficial for L3VPN Extension to use NVGRE and/or VXLAN as a data encapsulation format: o Maintain L3VPN implementation natively and decouple it completely from L2 overlay implementation. o Interwork with existing L3VPN customer o Seamlessly support L2 and L3 overlay interworking [E-IP-VPN] o BGP control plane works consistently with the data plane in term of multiple protocol support, i.e. the inner header on a data packet matches the address family being advertised by BGP route UPDATE message. o Save 14 bytes in each packet in a native L3 overlay and lower the probability of the packet fragmentation, and eliminate the unnecessary packet process Yong & Xu [Page 5] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 It is worth mention that the protocol extension in this document does not impact any mechanism that is specified in NVGRE [NVGRE] and VXLAN [VXLAN]. 3. Security Considerations The mechanism proposed in this document does not add any additional security concern beside what has been described in the NVGRE [NVGRE] and VXLAN [VXLAN]. 4. IANA Considerations The document does not require any IANA action. 5. References 5.1. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC2119, March 1997. [RFC2890] Dommety, G., "key and sequence Number Extentions to GRE", RFC2890, September 2000 [RFC4364] Rosen, E and Y. Rekhter, "BGP/MPLS IP Virtual Private Networks (VPNs)", RFC4364, February 2006. [RFC5512] Mohapatra, P., Rosen, E., "BGP Encapsulation Subsequent Address Family Identifiers (SAFI) and the BGP Tunnel Encapsulation Attribute", RFC5512, April 2009 5.2. Informative References [EVPN] Sajassi, A., EVPN, J., etc, "A Network Virtualization Overlay Solution using E-VPN", draft-sd-l2vpn-evpn- overlay-01, work in progress [DRAO] Rao, D., etc, "Layer-3 virtual network overlays based on BGP Layer-3 VPNs", draft-drao-bgp-l3vpn-virtual-network- overlays-01, work in progress [E-IP-VPN] Sajassi, A., etc, "E-VPN seamless Interoperability with Ip-VPN", draft-sajassi-l2vpn-evpn-ipvpn-interop-01, work in progress Yong & Xu [Page 6] Internet-Draft NVGRE/VXLAN for L3VPN Extension May 2013 [NVO3FRWK] Lasserre, M., etc, "Framework for DC Network Virtualization", draft-ietf-nv03-framework-02.txt, work in progress. [NVGRE] Sridharan, M., etc, "NVGRE: Network Virtualization using Generic Routing Encapsulation", draft-sridharan- virtualization-nvgre-02, work in progress [RFC3232] Raynolds, J., "ASSIGNED NUMBERS", RFC3232, January 2002 [VAN] VAN de Velde, G., etc, "BGP Remote-Next-Hop", draft- vandevelde-idr-remote-next-hop-03, work in progress. [VXLAN] Mahalingam, M., Dutt, D., etc, "VXLAN: A Framework for Overlaying Virtualized Layer 2 Networks over Layer 3 Networks", draft-mahalingam-dutt-dcops-vxlan-04.txt, work in progress Authors' Addresses Lucy Yong Huawei USA 5340 Legacy Drive Plano, TX 75025 U.S.A Phone: 469-277-5837 Email: lucy.yong@huawei.com Xiaohu Xu Huawei Technologies, Beijing, China Phone: +86-10-60610041 Email: xuxiaohu@huawei.com Yong & Xu [Page 7]