--- 1/draft-ietf-nvo3-dataplane-requirements-01.txt 2013-11-12 06:14:26.704969742 -0800 +++ 2/draft-ietf-nvo3-dataplane-requirements-02.txt 2013-11-12 06:14:26.744970768 -0800 @@ -1,47 +1,47 @@ Internet Engineering Task Force Nabil Bitar Internet Draft Verizon Intended status: Informational - Expires: January 2014 Marc Lasserre + Expires: May 2014 Marc Lasserre Florin Balus Alcatel-Lucent Thomas Morin France Telecom Orange Lizhong Jin Bhumip Khasnabish ZTE - July 1, 2013 + November 12, 2013 NVO3 Data Plane Requirements - draft-ietf-nvo3-dataplane-requirements-01.txt + draft-ietf-nvo3-dataplane-requirements-02.txt Status of this Memo 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 January 1, 2013. + This Internet-Draft will expire on May 12, 2014. Copyright Notice Copyright (c) 2013 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 @@ -53,57 +53,57 @@ Abstract Several IETF drafts relate to the use of overlay networks to support large scale virtual data centers. This draft provides a list of data plane requirements for Network Virtualization over L3 (NVO3) that have to be addressed in solutions documents. Table of Contents - 1. Introduction................................................3 - 1.1. Conventions used in this document.......................3 - 1.2. General terminology.....................................3 - 2. Data Path Overview..........................................4 - 3. Data Plane Requirements......................................5 - 3.1. Virtual Access Points (VAPs)............................5 - 3.2. Virtual Network Instance (VNI)..........................5 - 3.2.1. L2 VNI...............................................5 - 3.2.2. L3 VNI...............................................6 - 3.3. Overlay Module.........................................7 - 3.3.1. NVO3 overlay header...................................8 - 3.3.1.1. Virtual Network Context Identification..............8 - 3.3.1.2. Service QoS identifier..............................8 - 3.3.2. Tunneling function....................................9 - 3.3.2.1. LAG and ECMP.......................................10 - 3.3.2.2. DiffServ and ECN marking...........................10 - 3.3.2.3. Handling of BUM traffic............................11 - 3.4. External NVO3 connectivity.............................11 - 3.4.1. GW Types............................................12 - 3.4.1.1. VPN and Internet GWs...............................12 - 3.4.1.2. Inter-DC GW........................................12 - 3.4.1.3. Intra-DC gateways..................................12 - 3.4.2. Path optimality between NVEs and Gateways............12 - 3.4.2.1. Triangular Routing Issues (Traffic Tromboning)......13 - 3.5. Path MTU..............................................14 - 3.6. Hierarchical NVE.......................................15 - 3.7. NVE Multi-Homing Requirements..........................15 - 3.8. OAM...................................................16 - 3.9. Other considerations...................................16 - 3.9.1. Data Plane Optimizations.............................16 - 3.9.2. NVE location trade-offs..............................17 - 4. Security Considerations.....................................17 - 5. IANA Considerations........................................17 - 6. References.................................................18 - 6.1. Normative References...................................18 - 6.2. Informative References.................................18 - 7. Acknowledgments............................................19 + 1. Introduction..................................................3 + 1.1. Conventions used in this document........................3 + 1.2. General terminology......................................3 + 2. Data Path Overview............................................4 + 3. Data Plane Requirements.......................................5 + 3.1. Virtual Access Points (VAPs).............................5 + 3.2. Virtual Network Instance (VNI)...........................5 + 3.2.1. L2 VNI.................................................5 + 3.2.2. L3 VNI.................................................6 + 3.3. Overlay Module...........................................7 + 3.3.1. NVO3 overlay header....................................8 + 3.3.1.1. Virtual Network Context Identification...............8 + 3.3.1.2. Service QoS identifier...............................8 + 3.3.2. Tunneling function.....................................9 + 3.3.2.1. LAG and ECMP........................................10 + 3.3.2.2. DiffServ and ECN marking............................10 + 3.3.2.3. Handling of BUM traffic.............................11 + 3.4. External NVO3 connectivity..............................11 + 3.4.1. GW Types..............................................12 + 3.4.1.1. VPN and Internet GWs................................12 + 3.4.1.2. Inter-DC GW.........................................12 + 3.4.1.3. Intra-DC gateways...................................12 + 3.4.2. Path optimality between NVEs and Gateways.............12 + 3.4.2.1. Load-balancing......................................14 + 3.4.2.2. Triangular Routing Issues (a.k.a. Traffic Tromboning)14 + 3.5. Path MTU................................................14 + 3.6. Hierarchical NVE........................................15 + 3.7. NVE Multi-Homing Requirements...........................15 + 3.8. Other considerations....................................16 + 3.8.1. Data Plane Optimizations..............................16 + 3.8.2. NVE location trade-offs...............................16 + 4. Security Considerations......................................17 + 5. IANA Considerations..........................................17 + 6. References...................................................17 + 6.1. Normative References....................................17 + 6.2. Informative References..................................17 + 7. Acknowledgments..............................................18 1. Introduction 1.1. 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]. In this document, these words will appear with that interpretation @@ -203,50 +203,49 @@ tenants. There are different VNI types differentiated by the virtual network service they provide to Tenant Systems. Network virtualization can be provided by L2 and/or L3 VNIs. 3.2.1. L2 VNI An L2 VNI MUST provide an emulated Ethernet multipoint service as if Tenant Systems are interconnected by a bridge (but instead by using - a set of NVO3 tunnels). The emulated bridge MAY be 802.1Q enabled + a set of NVO3 tunnels). The emulated bridge could be 802.1Q enabled (allowing use of VLAN tags as a VAP). An L2 VNI provides per tenant virtual switching instance with MAC addressing isolation and L3 tunneling. Loop avoidance capability MUST be provided. Forwarding table entries provide mapping information between tenant system MAC addresses and VAPs on directly connected VNIs and L3 - tunnel destination addresses over the overlay. Such entries MAY be + tunnel destination addresses over the overlay. Such entries could be populated by a control or management plane, or via data plane. - In the absence of a management or control plane, data plane learning - MUST be used to populate forwarding tables. As frames arrive from - VAPs or from overlay tunnels, standard MAC learning procedures are - used: The tenant system source MAC address is learned against the - VAP or the NVO3 tunneling encapsulation source address on which the - frame arrived. This implies that unknown unicast traffic be flooded - i.e. broadcast. + By default, data plane learning MUST be used to populate forwarding + tables. As frames arrive from VAPs or from overlay tunnels, standard + MAC learning procedures are used: The tenant system source MAC + address is learned against the VAP or the NVO3 tunneling + encapsulation source address on which the frame arrived. This + implies that unknown unicast traffic will be flooded (i.e. + broadcast). When flooding is required, either to deliver unknown unicast, or broadcast or multicast traffic, the NVE MUST either support ingress - replication or multicast. In this latter case, the NVE MUST have one - or more multicast trees that can be used by local VNIs for flooding - to NVEs belonging to the same VN. For each VNI, there is one - flooding tree, and a multicast tree may be dedicated per VNI or - shared across VNIs. In such cases, multiple VNIs MAY share the same - default flooding tree. The flooding tree is equivalent with a + replication or multicast. + + When using multicast, the NVE MUST have one or more multicast trees + that can be used by local VNIs for flooding to NVEs belonging to the + same VN. For each VNI, there is at least one flooding tree used for + Broadcast, Unknown Unicast and Multicast forwarding. This tree MAY + be shared across VNIs. The flooding tree is equivalent with a multicast (*,G) construct where all the NVEs for which the - corresponding VNI is instantiated are members. The multicast tree - MAY be established automatically via routing and signaling or pre- - provisioned. + corresponding VNI is instantiated are members. When tenant multicast is supported, it SHOULD also be possible to select whether the NVE provides optimized multicast trees inside the VNI for individual tenant multicast groups or whether the default VNI flooding tree is used. If the former option is selected the VNI SHOULD be able to snoop IGMP/MLD messages in order to efficiently join/prune Tenant System from multicast trees. 3.2.2. L3 VNI @@ -268,27 +267,25 @@ L2 and L3 VNIs can be deployed in isolation or in combination to optimize traffic flows per tenant across the overlay network. For example, an L2 VNI may be configured across a number of NVEs to offer L2 multi-point service connectivity while a L3 VNI can be co- located to offer local routing capabilities and gateway functionality. In addition, integrated routing and bridging per tenant MAY be supported on an NVE. An instantiation of such service may be realized by interconnecting an L2 VNI as access to an L3 VNI on the NVE. - The L3 VNI does not require support for Broadcast and Unknown - Unicast traffic. The L3 VNI MAY provide support for customer - multicast groups. When multicast is supported, it SHOULD be possible - to select whether the NVE provides optimized multicast trees inside - the VNI for individual tenant multicast groups or whether a default - VNI multicasting tree, where all the NVEs of the corresponding VNI - are members, is used. + When multicast is supported, it MAY be possible to select whether + the NVE provides optimized multicast trees inside the VNI for + individual tenant multicast groups or whether a default VNI + multicasting tree, where all the NVEs of the corresponding VNI are + members, is used. 3.3. Overlay Module The overlay module performs a number of functions related to NVO3 header and tunnel processing. The following figure shows a generic NVO3 encapsulated frame: +--------------------------+ | Tenant Frame | @@ -310,42 +306,41 @@ this packet. . Outer underlay header: Can be either IP or MPLS . Outer link layer header: Header specific to the physical transmission link used 3.3.1. NVO3 overlay header An NVO3 overlay header MUST be included after the underlay tunnel - header when forwarding tenant traffic. Note that this information - can be carried within existing protocol headers (when overloading of - specific fields is possible) or within a separate header. + header when forwarding tenant traffic. + + Note that this information can be carried within existing protocol + headers (when overloading of specific fields is possible) or within + a separate header. 3.3.1.1. Virtual Network Context Identification The overlay encapsulation header MUST contain a field which allows the encapsulated frame to be delivered to the appropriate virtual - network endpoint by the egress NVE. The egress NVE uses this field - to determine the appropriate virtual network context in which to - process the packet. This field MAY be an explicit, unique (to the - administrative domain) virtual network identifier (VNID) or MAY - express the necessary context information in other ways (e.g. a - locally significant identifier). + network endpoint by the egress NVE. - It SHOULD be aligned on a 32-bit boundary so as to make it - efficiently processable by the data path. It MUST be distributable - by a control-plane or configured via a management plane. + The egress NVE uses this field to determine the appropriate virtual + network context in which to process the packet. This field MAY be an + explicit, unique (to the administrative domain) virtual network + identifier (VNID) or MAY express the necessary context information + in other ways (e.g. a locally significant identifier). In the case of a global identifier, this field MUST be large enough to scale to 100's of thousands of virtual networks. Note that there - is no such constraint when using a local identifier. + is typically no such constraint when using a local identifier. 3.3.1.2. Service QoS identifier Traffic flows originating from different applications could rely on differentiated forwarding treatment to meet end-to-end availability and performance objectives. Such applications may span across one or more overlay networks. To enable such treatment, support for multiple Classes of Service across or between overlay networks MAY be required. @@ -388,91 +383,95 @@ ISID tags and MPLS TC bits in the VPN labels. 3.3.2. Tunneling function This section describes the underlay tunneling requirements. From an encapsulation perspective, IPv4 or IPv6 MUST be supported, both IPv4 and IPv6 SHOULD be supported, MPLS tunneling MAY be supported. 3.3.2.1. LAG and ECMP - For performance reasons, multipath over LAG and ECMP paths SHOULD be + For performance reasons, multipath over LAG and ECMP paths MAY be supported. LAG (Link Aggregation Group) [IEEE 802.1AX-2008] and ECMP (Equal Cost Multi Path) are commonly used techniques to perform load- balancing of microflows over a set of a parallel links either at Layer-2 (LAG) or Layer-3 (ECMP). Existing deployed hardware implementations of LAG and ECMP uses a hash of various fields in the encapsulation (outermost) header(s) (e.g. source and destination MAC addresses for non-IP traffic, source and destination IP addresses, L4 protocol, L4 source and destination port numbers, etc). Furthermore, hardware deployed for the underlay network(s) will be most often unaware of the carried, innermost L2 frames or L3 packets - transmitted by the TS. Thus, in order to perform fine-grained load- - balancing over LAG and ECMP paths in the underlying network, the - encapsulation MUST result in sufficient entropy to exercise all - paths through several LAG/ECMP hops. The entropy information MAY be - inferred from the NVO3 overlay header or underlay header. If the - overlay protocol does not support the necessary entropy information - or the switches/routers in the underlay do not support parsing of - the additional entropy information in the overlay header, underlay - switches and routers should be programmable, i.e. select the - appropriate fields in the underlay header for hash calculation based - on the type of overlay header. + transmitted by the TS. + + Thus, in order to perform fine-grained load-balancing over LAG and + ECMP paths in the underlying network, the encapsulation MUST result + in sufficient entropy to exercise all paths through several LAG/ECMP + hops. + + The entropy information can be inferred from the NVO3 overlay header + or underlay header. If the overlay protocol does not support the + necessary entropy information or the switches/routers in the + underlay do not support parsing of the additional entropy + information in the overlay header, underlay switches and routers + should be programmable, i.e. select the appropriate fields in the + underlay header for hash calculation based on the type of overlay + header. All packets that belong to a specific flow MUST follow the same path in order to prevent packet re-ordering. This is typically achieved by ensuring that the fields used for hashing are identical for a given flow. - All paths available to the overlay network SHOULD be used - efficiently. Different flows SHOULD be distributed as evenly as + The goal is for all paths available to the overlay network to be + used efficiently. Different flows should be distributed as evenly as possible across multiple underlay network paths. For instance, this can be achieved by ensuring that some fields used for hashing are randomly generated. 3.3.2.2. DiffServ and ECN marking When traffic is encapsulated in a tunnel header, there are numerous options as to how the Diffserv Code-Point (DSCP) and Explicit Congestion Notification (ECN) markings are set in the outer header and propagated to the inner header on decapsulation. [RFC2983] defines two modes for mapping the DSCP markings from inner to outer headers and vice versa. The Uniform model copies the inner DSCP marking to the outer header on tunnel ingress, and copies that outer header value back to the inner header at tunnel egress. The Pipe model sets the DSCP value to some value based on local policy at ingress and does not modify the inner header on egress. Both models SHOULD be supported. - ECN marking MUST be performed according to [RFC6040] which describes - the correct ECN behavior for IP tunnels. + [RFC6040] defines ECN marking and processing for IP tunnels. 3.3.2.3. Handling of BUM traffic NVO3 data plane support for either ingress replication or point-to- multipoint tunnels is required to send traffic destined to multiple locations on a per-VNI basis (e.g. L2/L3 multicast traffic, L2 broadcast and unknown unicast traffic). It is possible that both methods be used simultaneously. There is a bandwidth vs state trade-off between the two approaches. - User-definable knobs MUST be provided to select which method(s) gets - used based upon the amount of replication required (i.e. the number - of hosts per group), the amount of multicast state to maintain, the - duration of multicast flows and the scalability of multicast - protocols. + User-configurable knobs MUST be provided to select which method(s) + gets used based upon the amount of replication required (i.e. the + number of hosts per group), the amount of multicast state to + maintain, the duration of multicast flows and the scalability of + multicast protocols. - When ingress replication is used, NVEs MUST track for each VNI the - related tunnel endpoints to which it needs to replicate the frame. + When ingress replication is used, NVEs MUST maintain for each VNI + the related tunnel endpoints to which it needs to replicate the + frame. For point-to-multipoint tunnels, the bandwidth efficiency is increased at the cost of more state in the Core nodes. The ability to auto-discover or pre-provision the mapping between VNI multicast trees to related tunnel endpoints at the NVE and/or throughout the core SHOULD be supported. 3.4. External NVO3 connectivity NVO3 services MUST interoperate with current VPN and Internet @@ -514,21 +513,21 @@ 3.4.1.3. Intra-DC gateways Even within one DC there may be End Devices that do not support NVO3 encapsulation, for example bare metal servers, hardware appliances and storage. A gateway device, e.g. a ToR, is required to translate the NVO3 to Ethernet VLAN encapsulation. 3.4.2. Path optimality between NVEs and Gateways - Within the NVO3 overlay, a default assumption is that NVO3 traffic + Within an NVO3 overlay, a default assumption is that NVO3 traffic will be equally load-balanced across the underlying network consisting of LAG and/or ECMP paths. This assumption is valid only as long as: a) all traffic is load-balanced equally among each of the component-links and paths; and, b) each of the component- links/paths is of identical capacity. During the course of normal operation of the underlying network, it is possible that one, or more, of the component-links/paths of a LAG may be taken out-of- service in order to be repaired, e.g.: due to hardware failure of cabling, optics, etc. In such cases, the administrator should configure the underlying network such that an entire LAG bundle in @@ -555,64 +554,71 @@ On the other hand, for Inter-DC and DC to External Network cases that use a WAN, the costs of the underlying network and/or service (e.g.: IPVPN service) are more expensive; therefore, there is a requirement on administrators to both: a) ensure high availability (active-backup failover or active-active load-balancing); and, b) maintaining substantial utilization of the WAN transport capacity at nearly all times, particularly in the case of active-active load- balancing. With respect to the dataplane requirements of NVO3 solutions, in the case of active-backup fail-over, all of the - ingress NVE's MUST dynamically adapt to the failure of an active NVE - GW when the backup NVE GW announces itself into the NVO3 overlay + ingress NVE's need to dynamically adapt to the failure of an active + NVE GW when the backup NVE GW announces itself into the NVO3 overlay immediately following a failure of the previously active NVE GW and update their forwarding tables accordingly, (e.g.: perhaps through dataplane learning and/or translation of a gratuitous ARP, IPv6 - Router Advertisement, etc.) Note that active-backup fail-over could - be used to accomplish a crude form of load-balancing by, for - example, manually configuring each tenant to use a different NVE GW, - in a round-robin fashion. On the other hand, with respect to active- - active load-balancing across physically separate NVE GW's (e.g.: - two, separate chassis) an NVO3 solution SHOULD support forwarding - tables that can simultaneously map a single egress NVE to more than - one NVO3 tunnels. The granularity of such mappings, in both active- - backup and active-active, MUST be unique to each tenant. + Router Advertisement). Note that active-backup fail-over could be + used to accomplish a crude form of load-balancing by, for example, + manually configuring each tenant to use a different NVE GW, in a + round-robin fashion. - 3.4.2.1. Triangular Routing Issues (Traffic Tromboning) + 3.4.2.1. Load-balancing + + When using active-active load-balancing across physically separate + NVE GW's (e.g.: two, separate chassis) an NVO3 solution SHOULD + support forwarding tables that can simultaneously map a single + egress NVE to more than one NVO3 tunnels. The granularity of such + mappings, in both active-backup and active-active, MUST be specific + to each tenant. + + 3.4.2.2. Triangular Routing Issues (a.k.a. Traffic Tromboning) L2/ELAN over NVO3 service may span multiple racks distributed across different DC regions. Multiple ELANs belonging to one tenant may be interconnected or connected to the outside world through multiple Router/VRF gateways distributed throughout the DC regions. In this scenario, without aid from an NVO3 or other type of solution, traffic from an ingress NVE destined to External gateways will take a non-optimal path that will result in higher latency and costs, (since it is using more expensive resources of a WAN). In the case of traffic from an IP/MPLS network destined toward the entrance to an NVO3 overlay, well-known IP routing techniques MAY be used to optimize traffic into the NVO3 overlay, (at the expense of additional routes in the IP/MPLS network). In summary, these issues are well known as triangular routing. Procedures for gateway selection to avoid triangular routing issues - SHOULD be provided. The details of such procedures are, most likely, - part of the NVO3 Management and/or Control Plane requirements and, - thus, out of scope of this document. However, a key requirement on - the dataplane of any NVO3 solution to avoid triangular routing is - stated above, in Section 3.4.2, with respect to active-active load- - balancing. More specifically, an NVO3 solution SHOULD support - forwarding tables that can simultaneously map a single egress NVE to - more than one NVO3 tunnels. The expectation is that, through the - Control and/or Management Planes, this mapping information MAY be - dynamically manipulated to, for example, provide the closest - geographic and/or topological exit point (egress NVE) for each - ingress NVE. + SHOULD be provided. + + The details of such procedures are, most likely, part of the NVO3 + Management and/or Control Plane requirements and, thus, out of scope + of this document. However, a key requirement on the dataplane of any + NVO3 solution to avoid triangular routing is stated above, in + Section 3.4.2, with respect to active-active load-balancing. More + specifically, an NVO3 solution SHOULD support forwarding tables that + can simultaneously map a single egress NVE to more than one NVO3 + tunnel. + + The expectation is that, through the Control and/or Management + Planes, this mapping information may be dynamically manipulated to, + for example, provide the closest geographic and/or topological exit + point (egress NVE) for each ingress NVE. 3.5. Path MTU The tunnel overlay header can cause the MTU of the path to the egress tunnel endpoint to be exceeded. IP fragmentation SHOULD be avoided for performance reasons. The interface MTU as seen by a Tenant System SHOULD be adjusted such that no fragmentation is needed. This can be achieved by @@ -630,21 +636,21 @@ o The underlay network MAY be designed in such a way that the MTU can accommodate the extra tunnel overhead. 3.6. Hierarchical NVE It might be desirable to support the concept of hierarchical NVEs, such as spoke NVEs and hub NVEs, in order to address possible NVE performance limitations and service connectivity optimizations. - For instance, spoke NVE functionality MAY be used when processing + For instance, spoke NVE functionality may be used when processing capabilities are limited. A hub NVE would provide additional data processing capabilities such as packet replication. NVEs can be either connected in an any-to-any or hub and spoke topology on a per VNI basis. 3.7. NVE Multi-Homing Requirements Multi-homing techniques SHOULD be used to increase the reliability of an nvo3 network. It is also important to ensure that physical @@ -666,67 +672,38 @@ system is co-located with an NVE, IP routing can be relied upon to handle routing over diverse links to TORs. External connectivity MAY be handled by two or more nvo3 gateways. Each gateway is connected to a different domain (e.g. ISP) and runs BGP multi-homing. They serve as an access point to external networks such as VPNs or the Internet. When a connection to an upstream router is lost, the alternative connection is used and the failed route withdrawn. - 3.8. OAM - - NVE MAY be able to originate/terminate OAM messages for connectivity - verification, performance monitoring, statistic gathering and fault - isolation. Depending on configuration, NVEs SHOULD be able to - process or transparently tunnel OAM messages, as well as supporting - alarm propagation capabilities. - - Given the critical requirement to load-balance NVO3 encapsulated - packets over LAG and ECMP paths, it will be equally critical to - ensure existing and/or new OAM tools allow NVE administrators to - proactively and/or reactively monitor the health of various - component-links that comprise both LAG and ECMP paths carrying NVO3 - encapsulated packets. For example, it will be important that such - OAM tools allow NVE administrators to reveal the set of underlying - network hops (topology) in order that the underlying network - administrators can use this information to quickly perform fault - isolation and restore the underlying network. - - The NVE MUST provide the ability to reveal the set of ECMP and/or - LAG paths used by NVO3 encapsulated packets in the underlying - network from an ingress NVE to egress NVE. The NVE MUST provide the - ability to provide a "ping"-like functionality that can be used to - determine the health (liveness) of remote NVE's or their VNI's. The - NVE SHOULD provide a "ping"-like functionality to more expeditiously - aid in troubleshooting performance problems, i.e.: blackholing or - other types of congestion occurring in the underlying network, for - NVO3 encapsulated packets carried over LAG and/or ECMP paths. - - 3.9. Other considerations + 3.8. Other considerations - 3.9.1. Data Plane Optimizations + 3.8.1. Data Plane Optimizations Data plane forwarding and encapsulation choices SHOULD consider the limitation of possible NVE implementations, specifically in software based implementations (e.g. servers running VSwitches) NVE SHOULD provide efficient processing of traffic. For instance, packet alignment, the use of offsets to minimize header parsing, padding techniques SHOULD be considered when designing NVO3 encapsulation types. The NV03 encapsulation/decapsulation processing in software-based NVEs SHOULD make use of hardware assist provided by NICs in order to speed up packet processing. - 3.9.2. NVE location trade-offs + 3.8.2. NVE location trade-offs In the case of DC traffic, traffic originated from a VM is native Ethernet traffic. This traffic can be switched by a local VM switch or ToR switch and then by a DC gateway. The NVE function can be embedded within any of these elements. The NVE function can be supported in various DC network elements such as a VM, VM switch, ToR switch or DC GW. The following criteria SHOULD be considered when deciding where the @@ -808,25 +785,22 @@ [RFC6391] Bryant, S. et al, "Flow-Aware Transport of Pseudowires over an MPLS Packet Switched Network", RFC6391, November 2011 7. Acknowledgments In addition to the authors the following people have contributed to this document: - Shane Amante, Level3 - - Dimitrios Stiliadis, Rotem Salomonovitch, Alcatel-Lucent - - Larry Kreeger, Cisco + Shane Amante, Dimitrios Stiliadis, Rotem Salomonovitch, Larry + Kreeger, and Eric Gray. This document was prepared using 2-Word-v2.0.template.dot. Authors' Addresses Nabil Bitar Verizon 40 Sylvan Road Waltham, MA 02145 Email: nabil.bitar@verizon.com