idnits 2.17.1 draft-yizhou-nvo3-hpvr2nve-cp-req-00.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 == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'SHOULD not' in this paragraph: External NVE connecting to destination hypervisor 2 has to associate the migrating VM with it by saving VM's MAC and/or IP addresses, its VN, locally significant VID if any, and provisioning other network related parameters of VM. The NVE may be informed about the VM's peer VMs, storage devices and other network appliances with which the VM needs to communicate or is communicating. VM on destination hypervisor 2 SHOULD not go to running state before all the network provisioning and binding has been done. == Using lowercase 'not' together with uppercase 'MUST', 'SHALL', 'SHOULD', or 'RECOMMENDED' is not an accepted usage according to RFC 2119. Please use uppercase 'NOT' together with RFC 2119 keywords (if that is what you mean). Found 'SHOULD not' in this paragraph: VM on source hypervisor and destination hypervisor SHOULD not be in running state at the same time during migration. VM on source hypervisor goes into shutdown state only when VM on destination hypervisor has successfully been entering the running state. It is possible that VM on the source hypervisor stays in migrating state for a while after VM on the destination hypervisor is in running state. -- The document date (May 16, 2014) is 3631 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'I-D.ietf-nvo3-framework' is defined on line 667, but no explicit reference was found in the text == Unused Reference: '8021Q' is defined on line 690, but no explicit reference was found in the text == Outdated reference: A later version (-09) exists of draft-ietf-nvo3-framework-05 == Outdated reference: A later version (-05) exists of draft-ietf-nvo3-nve-nva-cp-req-01 == Outdated reference: A later version (-04) exists of draft-ietf-opsawg-vmm-mib-00 Summary: 0 errors (**), 0 flaws (~~), 8 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 NVO3 Working Group Yizhou Li 3 INTERNET-DRAFT Lucy Yong 4 Intended Status: Informational Huawei Technologies 5 Lawrence Kreeger 6 Cisco 7 Thomas Narten 8 IBM 9 David Black 10 EMC 11 Expires: November 17, 2014 May 16, 2014 13 Hypervisor to NVE Control Plane Requirements 14 draft-yizhou-nvo3-hpvr2nve-cp-req-00 16 Abstract 18 This document describes the control plane protocol requirements when 19 NVE is not co-located with the hypervisor on a server. A control 20 plane protocol (or protocols) between a hypervisor and its associated 21 external NVE(s) is used for the hypervisor to populate its virtual 22 machines states to the NVE(s) for further handling. This document 23 illustrates the functionalities required by such control plane 24 signaling protocols and outlines the high level requirements to be 25 fulfiled. Virtual machine states and state transitioning are 26 summarized to help clarifying the needed requirements. 28 Status of this Memo 30 This Internet-Draft is submitted to IETF in full conformance with the 31 provisions of BCP 78 and BCP 79. 33 Internet-Drafts are working documents of the Internet Engineering 34 Task Force (IETF), its areas, and its working groups. Note that 35 other groups may also distribute working documents as 36 Internet-Drafts. 38 Internet-Drafts are draft documents valid for a maximum of six months 39 and may be updated, replaced, or obsoleted by other documents at any 40 time. It is inappropriate to use Internet-Drafts as reference 41 material or to cite them other than as "work in progress." 43 The list of current Internet-Drafts can be accessed at 44 http://www.ietf.org/1id-abstracts.html 45 The list of Internet-Draft Shadow Directories can be accessed at 46 http://www.ietf.org/shadow.html 48 Copyright and License Notice 50 Copyright (c) 2013 IETF Trust and the persons identified as the 51 document authors. All rights reserved. 53 This document is subject to BCP 78 and the IETF Trust's Legal 54 Provisions Relating to IETF Documents 55 (http://trustee.ietf.org/license-info) in effect on the date of 56 publication of this document. Please review these documents 57 carefully, as they describe your rights and restrictions with respect 58 to this document. Code Components extracted from this document must 59 include Simplified BSD License text as described in Section 4.e of 60 the Trust Legal Provisions and are provided without warranty as 61 described in the Simplified BSD License. 63 Table of Contents 65 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 66 1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . 3 67 1.2 Target Scenarios . . . . . . . . . . . . . . . . . . . . . 4 68 2. VM Lifecycle . . . . . . . . . . . . . . . . . . . . . . . . . 6 69 2.1 VM Creation . . . . . . . . . . . . . . . . . . . . . . . . 6 70 2.2 VM Live Migration . . . . . . . . . . . . . . . . . . . . . 7 71 2.3 VM termination . . . . . . . . . . . . . . . . . . . . . . . 7 72 2.4 VM Pause, suspension and resumption . . . . . . . . . . . . 8 73 3. Hypervisor-to-NVE Signaling protocol functionality . . . . . . 8 74 3.1 VN connect and disconnect . . . . . . . . . . . . . . . . . 8 75 3.2 TSI associate and activate . . . . . . . . . . . . . . . . . 10 76 3.3 TSI disassociate, deactivate and clear . . . . . . . . . . . 12 77 4. Hypervisor-to-NVE Signaling Protocol requirements . . . . . . . 13 78 5. Security Considerations . . . . . . . . . . . . . . . . . . . . 14 79 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 15 80 7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 15 81 8. References . . . . . . . . . . . . . . . . . . . . . . . . . . 15 82 8.1 Normative References . . . . . . . . . . . . . . . . . . . 15 83 8.2 Informative References . . . . . . . . . . . . . . . . . . 15 84 Appendix A. IEEE 802.1Qbg VDP Illustration . . . . . . . . . . . . 16 85 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 19 87 1. Introduction 89 This document describes the control plane protocol requirements when 90 NVE is not co-located with the hypervisor on a server. A control 91 plane protocol (or protocols) between a hypervisor and its associated 92 external NVE(s) is used for the hypervisor to populate its virtual 93 machines states to the NVE(s) for further handling. This protocol is 94 mentioned in NVO3 problem statement [I-D.ietf-nvo3-overlay-problem- 95 statement] as the third work item. When TS and NVE are on the 96 separate devices, we also call it split TS-NVE architecture and it is 97 the primary interest in this document. 99 Virtual machine states and state transitioning are summarized in this 100 document to illustrates the functionalities required by the control 101 plane signaling protocols between hypervisor and the external NVE. 102 Then the high level requirements to be fulfiled are outlined. 104 This document uses the term "hypervisor" throughout when describing 105 the scenario where NVE functionality is implemented on a separate 106 device from the "hypervisor" that contains a VM connected to a VN. 107 In this context, the term "hypervisor" is meant to cover any device 108 type where the NVE functionality is offloaded in this fashion, e.g., 109 a Network Service Appliance. 111 This document often uses the term "VM" and "Tenant System" (TS) 112 interchangeably, even though a VM is just one type of Tenant System 113 that may connect to a VN. For example, a service instance within a 114 Network Service Appliance may be another type of TS. When this 115 document uses the term VM, it will in most cases apply to other types 116 of TSs. 118 1.1 Terminology 120 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 121 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 122 document are to be interpreted as described in RFC 2119 [RFC2119]. 124 This document uses the same terminology as found in [I-D.ietf-nvo3- 125 framework] and [I-D.ietf-nvo3-nve-nva-cp-req]. This section defines 126 additional terminology used by this document. 128 VN Profile: Meta data associated with a VN that is used by an NVE 129 when ingressing/egressing packets to/from a specific VN. Meta data 130 could include such information as ACLs, QoS settings, etc. The VN 131 Profile contains parameters that apply to the VN as a whole. Control 132 protocols could use the VN ID or VN Name to obtain the VN Profile. 134 VSI: Virtual Station Interface. [IEEE 802.1Qbg] 135 VDP: VSI Discovery and Configuration Protocol [IEEE 802.1Qbg] 137 1.2 Target Scenarios 139 In split TS-NVE architecture, an external NVE can provide an offload 140 of the encapsulation / decapsulation function, network policy 141 enforcement, as well as the VN Overlay protocol overheads. This 142 offloading may provide performance improvements and/or resource 143 savings to the End Device (e.g. hypervisor) making use of the 144 external NVE. 146 The following figures give example scenarios where the Tenant System 147 and NVE are on different devices. 149 Hypervisor Access Switch 150 +------------------+ +-----+-------+ 151 | +--+ +-------+ | | | | 152 | |VM|---| | | VLAN | | | 153 | +--+ |Virtual|---------+ NVE | +--- Underlying 154 | +--+ |Switch | | Trunk | | | Network 155 | |VM|---| | | | | | 156 | +--+ +-------+ | | | | 157 +------------------+ +-----+-------+ 158 Hypervisor with an External NVE 160 Access 161 Hypervisor Switch NVE 162 +------------------+ +-----+ +-----+ 163 | +--+ +-------+ | | | | | 164 | |VM|---| | | VLAN | | VLAN | | 165 | +--+ |Virtual|---------+ +-------+ +--- Underlying 166 | +--+ |Switch | | Trunk | | Trunk | | Network 167 | |VM|---| | | | | | | 168 | +--+ +-------+ | | | | | 169 +------------------+ +-----+ +-----+ 170 Hypervisor with an External NVE across an Ethernet Access Switch 171 Network Service Appliance Access Switch 172 +--------------------------+ +-----+-------+ 173 | +------------+ |\ | | | | 174 | |Net Service |----| \ | | | | 175 | |Instance | | \ | VLAN | | | 176 | +------------+ | |---------+ NVE | +--- Underlying 177 | +------------+ | | | Trunk| | | Network 178 | |Net Service |----| / | | | | 179 | |Instance | | / | | | | 180 | +------------+ |/ | | | | 181 +--------------------------+ +-----+-------+ 182 Physical Network Service Appliance with an External NVE 184 Figure 1 Split TS-NVE Architecture 186 Tenant Systems connect to NVEs via a Tenant System Interface (TSI). 187 The TSI logically connects to the NVE via a Virtual Access Point 188 (VAP) [I-D.ietf-nvo3-arch]. NVE may provide Layer 2 or Layer 3 189 forwarding. In split TS-NVE architecture, external NVE may be able to 190 reach multiple MAC and IP addresses via a TSI. For example, Tenant 191 Systems that are providing network services (such as firewall, load 192 balancer, VPN gateway) are likely to have complex address hierarchy. 193 It implies if a given TSI disassociates from one VN, all the MAC and 194 IP addresses are also disassociated. There is no need to signal the 195 deletion of every MAC or IP when the TSI is brought down or deleted. 196 In the majority of cases, a VM will be acting as a simple host that 197 will have a single TSI and single MAC and IP visible to the external 198 NVE. 200 1.3 Motivations and Purpose 202 The problem statement [I-D.ietf-nvo3-overlay-problem-statement], 203 discusses the needs for a control plane protocol (or protocols) to 204 populate each NVE with the state needed to perform its functions. 206 In one common scenario, an NVE provides overlay 207 encapsulation/decapsulation packet forwarding services to Tenant 208 Systems (TSs) that are co-resident with the NVE on the same End 209 Device (e.g. when the NVE is embedded within a hypervisor or a 210 Network Service Appliance). In such cases, there is no need for a 211 standardized protocol between the hypervisor and NVE, as the 212 interaction is implemented via software on a single device. While in 213 the split TS-NVE architecture scenarios, as shown in figure 1, some 214 control plane signaling protocol needs to run between hypervisor and 215 external NVE to pass the relevant state information. Such interaction 216 is mandatory. This document will identify the requirements for such 217 signaling protocol. 219 Section 2 describes VM states and state transitioning in its 220 lifecycle. Section 3 introduces Hypervisor-to-NVE signaling protocol 221 functionality derived from VM operations and network events. Section 222 4 outlines the requirements of the control plane protocol to achieve 223 the required functionality. 225 2. VM Lifecycle 227 [I-D.ietf-opsawg-vmm-mib] shows the state transition of a VM in its 228 figure 2. Some of the VM states are of the interest to the external 229 NVE. This section illustrates the relevant phases or event in VM 230 lifecycle. It should be noted that the following subsections do not 231 give an exhaustive traversal of VM lifecycle state. They are intended 232 as the illustrative examples which are relevant to split TS-NVE 233 architecture, not as prescriptive text; the goal is to capture 234 sufficient detail to set a context for the signaling protocol 235 functionality and requirements described in the following sections. 237 2.1 VM Creation 239 VM creation runs through the states in the order of preparing, 240 shutdown and running [I-D.ietf-opsawg-vmm-mib]. The end device 241 allocates and initializes local virtual resources like storage in the 242 VM preparing state. In shutdown state, VM has everything ready except 243 that CPU execution is not scheduled by the hypervisor and VM's memory 244 is not resident in the hypervisor. From the shutdown state to running 245 state, normally it requires the human execution or system triggered 246 event. Running state indicates the VM is in the normal execution 247 state. Frame can be sent and received correctly. No ongoing 248 migration, suspension or shutdown is in process. 250 In VM creation phase, tenant system has to be associated with the 251 external NVE. Association here indicates that hypervisor and the 252 external NVE have signaled each other and reached some agreement. 253 Relevant parameters or information have been provisioned properly. 254 External NVE should be informed with VM's MAC address and/or IP 255 address. Another example is that hypervisor may use a locally 256 significant VLAN ID to indicate the traffic destined to a specified 257 VN. Both hypervisor and NVE sides should agree on that VID value for 258 later traffic identification and forwarding. 260 External NVE needs to do some preparation work before it signals 261 successful association with tenant system. Such preparation work may 262 include locally saving the states and binding information of the 263 tenant system and its VN, communicating with peer NVEs and/or NVA for 264 network provisioning, etc. 266 Tenant System association should be performed before VM enters 267 running state, preferably in shutdown state. If association with 268 external NVE fails, VM should not go into running state. 270 2.2 VM Live Migration 272 Live migration is sometimes referred to as "hot" migration, in that 273 from an external viewpoint, the VM appears to continue to run while 274 being migrated to another server (e.g., TCP connections generally 275 survive this class of migration). In contrast, suspend/resume (or 276 "cold") migration consists of suspending VM execution on one server 277 and resuming it on another. For simplicity, the following abstract 278 summary about live migration assumes shared storage, so that the VM's 279 storage is accessible to the source and destination servers. Assume 280 VM migrates from hypervisor 1 to hypervisor 2. VM live migration 281 involves the state transition on both hypervisors, source hypervisor 282 1 and destination hypervisor 2. VM state on source hypervisor 1 283 transits from running to migrating and then to shutdown [I-D.ietf- 284 opsawg-vmm-mib]. VM state on destination hypervisor 2 transits from 285 shutdown to migrating and then running. 287 External NVE connecting to destination hypervisor 2 has to associate 288 the migrating VM with it by saving VM's MAC and/or IP addresses, its 289 VN, locally significant VID if any, and provisioning other network 290 related parameters of VM. The NVE may be informed about the VM's peer 291 VMs, storage devices and other network appliances with which the VM 292 needs to communicate or is communicating. VM on destination 293 hypervisor 2 SHOULD not go to running state before all the network 294 provisioning and binding has been done. 296 VM on source hypervisor and destination hypervisor SHOULD not be in 297 running state at the same time during migration. VM on source 298 hypervisor goes into shutdown state only when VM on destination 299 hypervisor has successfully been entering the running state. It is 300 possible that VM on the source hypervisor stays in migrating state 301 for a while after VM on the destination hypervisor is in running 302 state. 304 2.3 VM termination 306 VM termination is also referred to as "powering off" a VM. VM 307 termination leads its state going to shutdown. There are two possible 308 causes to terminate a VM [I-D.ietf-opsawg-vmm-mib], one is the normal 309 "power off" of a running VM; the other is that VM has been migrated 310 to other place and the VM image on the source hypervisor has to stop 311 executing and to be shutdown. 313 In VM termination, the external NVE connecting to that VM needs to 314 deprovision the VM, i.e. delete the network parameters associated 315 with that VM. In other words, external NVE has to de-associate the 316 VM. 318 2.4 VM Pause, suspension and resumption 320 VM pause event leads VM transiting from running state to paused 321 state. Paused state indicates VM is resident in memory but no longer 322 scheduled to execute by the hypervisor [I-D.ietf-opsawg-vmm-mib]. VM 323 can be easily re-activated from paused state to running state. 325 VM suspension leads VM to transit state from running to suspended and 326 VM resumption leads VM to transit state from suspended to running. 327 Suspended state means the memory and CPU execution state of the 328 virtual machine are saved to persistent store. During this state, 329 the virtual machine is not scheduled to execute by the hypervisor [I- 330 D.ietf-opsawg-vmm-mib]. 332 In split TS-NVE architecture, external NVE should keep any paused or 333 suspended VM in association as VM can return to running state at any 334 time. 336 3. Hypervisor-to-NVE Signaling protocol functionality 338 3.1 VN connect and disconnect 340 When an NVE is external, a protocol is needed between the End Device 341 (e.g. Hypervisor) making use of the external NVE and the external NVE 342 in order to make the NVE aware of the changing VN membership 343 requirements of the Tenant Systems within the End Device. 345 A key driver for using a protocol rather than using static 346 configuration of the external NVE is because the VN connectivity 347 requirements can change frequently as VMs are brought up, moved and 348 brought down on various hypervisors throughout the data center. 350 +---------------+ Recv VN_connect; +-------------------+ 351 |VN_Disconnected| return Local_Tag value |VN_Connected | 352 +---------------+ for VN if successful; +-------------------+ 353 |VN_ID; |-------------------------->|VN_ID; | 354 |VN_State= | |VN_State=connected;| 355 |disconnected; | |Num_TSI_Associated;| 356 | |<----Recv VN_disconnect----|Local_Tag; | 357 +---------------+ |VN_Context; | 358 +-------------------+ 360 Figure 2 State machine of a VAP instance on an external NVE 362 Figure 2 show the state machine for a VAP on the external NVE. An NVE 363 that supports the hypervisor to NVE signaling protocol should support 364 one instance of the state machine for each active VN. The state 365 transition on the external NVE is normally triggered by the 366 hypervisor-facing side events and behaviors. Some of the interleaved 367 interaction between NVE and NVA will be illustrated for better 368 understanding of the whole procedures; while some of them may not be 369 shown. More detailed information regarding that is available in [I- 370 D.ietf-nvo3-nve-nva-cp-req]. 372 The NVE must be notified when an End Device requires connection to a 373 particular VN and when it no longer requires connection. In addition, 374 the external NVE must provide a local tag value for each connected VN 375 to the End Device to use for exchange of packets between the End 376 Device and the NVE (e.g. a locally significant 802.1Q tag value). How 377 "local" the significance is depends on whether the Hypervisor has a 378 direct physical connection to the NVE (in which case the significance 379 is local to the physical link), or whether there is an Ethernet 380 switch (e.g. a blade switch) connecting the Hypervisor to the NVE (in 381 which case the significance is local to the intervening switch and 382 all the links connected to it). 384 These VLAN tags are used to differentiate between different VNs as 385 packets cross the shared access network to the external NVE. When the 386 NVE receives packets, it uses the VLAN tag to identify the VN of 387 packets coming from a given TSI, strips the tag, and adds the 388 appropriate overlay encapsulation for that VN and send to the 389 corresponding VAP. 391 The Identification of the VN in this protocol could either be through 392 a VN Name or a VN ID. A globally unique VN Name facilitates 393 portability of a Tenant's Virtual Data Center. Once an NVE receives a 394 VN connect indication, the NVE needs a way to get a VN Context 395 allocated (or receive the already allocated VN Context) for a given 396 VN Name or ID (as well as any other information needed to transmit 397 encapsulated packets). How this is done is the subject of the NVE- 398 to-NVA (called NVE-to-NVA in this document) protocol which are part 399 of work items 1 and 2 in [I-D.ietf-nvo3-overlay-problem-statement]. 401 VN_connect message can be explicit or implicit. Explicit means the 402 hypervisor sending a message explicitly to request for the connection 403 to a VN. Implicit means the external NVE receives other messages, 404 e.g. very first TSI associate message for a given VN as in next 405 subsection, to implicitly indicate its interest to connect to a VN. 407 A VN_disconnect message will make NVE release all the resources for 408 that disconnected VN and transit to VN_disconnected state. The local 409 tag assigned for that VN can possibly be reclaimed by other VN. 411 3.2 TSI associate and activate 413 Typically, a TSI is assigned a single MAC address and all frames 414 transmitted and received on that TSI use that single MAC address. As 415 mentioned earlier, it is also possible for a Tenant System to 416 exchange frames using multiple MAC addresses or packets with multiple 417 IP addresses. 419 Particularly in the case of a TS that is forwarding frames or packets 420 from other TSs, the NVE will need to communicate the mapping between 421 the NVE's IP address (on the underlying network) and ALL the 422 addresses the TS is forwarding on behalf of to NVA in each 423 corresponding VN. 425 The NVE has two ways in which it can discover the tenant addresses 426 for which frames must be forwarded to a given End Device (and 427 ultimately to the TS within that End Device). 429 1. It can glean the addresses by inspecting the source addresses in 430 packets it receives from the End Device. 432 2. The hypervisor can explicitly signal the address associations of 433 a TSI to the external NVE. The address association includes all the 434 MAC and/or IP addresses possibly used as source addresses in a packet 435 sent from the hypervisor to external NVE. External NVE may further 436 use this information to filter the future traffic from the 437 hypervisor. 439 To perform the second approach above, the "hypervisor-to-NVE" 440 protocol requires a means to allow End Devices to communicate new 441 tenant addresses associations for a given TSI within a given VN. 443 Figure 3 shows the state machine for a TSI connecting to a VAP on the 444 external NVE. An NVE that supports the hypervisor to NVE signaling 445 protocol should support one instance of the state machine for each 446 TSI connecting to a given VN. 448 disassociate; +--------+ 449 +--------------->| Init |<--------clear-------+ 450 |or keepalive +--------+ | 451 |timer timeout; | | | 452 | | | | 453 | +--------+ | 454 | | | | 455 | associate | | activate | 456 | +-----------+ +-----------+ | 457 | | | | 458 | | | | 459 | \|/ \|/ | 460 +--------------------+ +---------------------+ 461 | Associated | | Activated | 462 +--------------------+ +---------------------+ 463 |TSI_ID; | |TSI_ID; | 464 |Port; |-----activate---->|Port; | 465 |VN_ID; | |VN_ID; | 466 |State=associated; | |State=activated ; |-+ 467 +-|Num_Of_Addr; |<---deactivate;---|Num_Of_Addr; | | 468 | |List_Of_Addr; | or keepactive List_Of_Addr; | | 469 | |ResetKeepaliveTimer;| timer timeout; |ResetKeepactiveTimer;| | 470 | +--------------------+ +---------------------+ | 471 | /|\ /|\ | 472 | | | | 473 +---------------------+ +-------------------+ 474 add/remove/updt addr; add/remove/updt addr; 475 or update port; or or update port; or 476 Recv keepalive pkt Recv keepactive pkt 477 from TSI; or data msg from TSI; 479 Figure 3 State machine of a TSI instance on an external NVE 481 Associated state of a TSI instance on an external NVE indicates all 482 the addresses for that TSI have already associated with the VAP of 483 the external NVE on port p for a given VN but no real traffic to and 484 from the TSI is expected and allowed to pass through. NVE has 485 reserved all the necessary resources for that TSI. NVE may report the 486 mappings of NVE's underlay IP address and the associated TSI 487 addresses to NVA and relevant network nodes may save such information 488 to its mapping table but not forwarding table. NVE may create ACL or 489 filter rules based on the associated TSI addresses on the attached 490 port p but not enable them yet. Local tag for the VN corresponding to 491 the TSI instance should be provisioned on port p to receive packets. 493 VM migration discussed section 2 may cause the hypervisor send 494 associate message to the NVE connecting the destination hypervisor 495 the VM migrates to. It is similar as the resource reservation request 496 to make sure the VM can be successfully migrated later. If such 497 association fails, VM may choose another destination hypervisor to 498 migrate to or alert with an administrative message. VM creation event 499 may also lead to the same practice. 501 Activated state of a TSI instance on an external NVE indicates that 502 all the addresses for that TSI functioning correctly on port p and 503 traffic can be received from and sent to that TSI on NVE. The 504 mappings of NVE's underlay IP address and the associated TSI 505 addresses should be put into the forwarding table rather than the 506 mapping table on relevant network nodes. ACL or filter rules based on 507 the associated TSI addresses on the attached port p in NVE are 508 enabled. Local tag for the VN corresponding to the TSI instance MUST 509 be provisioned on port p to receive packets. 511 Activate message makes the state transit from Init or Associated to 512 Activated. VM creation, VM migration and VM resumption events 513 discussed in section 4 may trigger activate message to be sent from 514 the hypervisor to the external NVE. 516 As mentioned in last subsection, associate or activate message from 517 the very first TSI connecting to a VN on an NVE is also considered as 518 the implicit VN_connect signal to create a VAP for that VN. 520 TSI information may get updated either in Associated or Activated 521 state. Add or remove the associated addresses, update current 522 associated addresses for example updating IP for a given MAC, update 523 NVE port information from which the message receives are all 524 considered as TSI information updating. Such update does not change 525 the state of TSI. When any address associated to a given TSI changes, 526 NVE should inform the NVA to update the mapping information on NVE's 527 underlying address and the associated TSI addresses. NVE should also 528 change its local ACL or filter settings accordingly for the relevant 529 addresses. Port information update will cause the local tag for the 530 VN corresponding to the TSI instance provisioned on new port p and 531 removed from old port. 533 NVE keeps a timer for each TSI instance associated or activated on 534 it. When NVE receives the keepalive or keepactive message for a TSI 535 instance, it should reset the timer. Keepactive timer may also be 536 reset by receiving the data packet from any associated address of the 537 corresponding TSI instance. Keepactive timer times out leads the 538 state transiting from Activated to Associated. Keepalive timer times 539 out leads the state transiting from Associated to Init. 541 3.3 TSI disassociate, deactivate and clear 543 Disassociate and deactivate conceptually are the reverse behaviors of 544 associate and activate. From Activated state to Associated state, NVE 545 needs to make sure the resources still reserved but the addresses 546 associated to the TSI not functioning and no traffic to and from the 547 TSI expected and allowed to pass through. For example, NVE needs to 548 inform NVA to remove the relevant addresses mapping information from 549 forwarding or routing table. ACL or filtering rules regarding the 550 relevant addresses should be disabled. From Associated or Activated 551 state to Init state, NVE will release all the resource relevant to 552 TSI instances. NVE should also inform the NVA to remove the relevant 553 entries from mapping table. ACL or filtering rules regarding the 554 relevant addresses should be removed. Local tag provisioning on the 555 connecting port on NVE should be cleared. 557 VM suspension discussed in section 2 may cause the relevant TSI 558 instance(s) on NVE transit from Activated to Associated state. VM 559 pause normally does not affect the state of the relevant TSI 560 instance(s) on NVE as the VM is expected to run again soon. VM 561 shutdown will cause the relevant TSI instance(s) on NVE transit to 562 Init state from Activated state. All resources should be released. 564 VM migration will lead the TSI instance on the source NVE to leave 565 Activated state. Such state transition on source NVE should not occur 566 earlier than the TSI instance on the destination NVE transits to 567 Activated state. Otherwise traffic interruption may occur. When a VM 568 migrates to another hypervisor connecting to the same NVE, i.e. 569 source and destination NVE are the same, NVE should use TSI_ID and 570 incoming port to differentiate two TSI instance. 572 Although the triggering messages for state transition shown in Figure 573 3 does not indicate the difference between VM creation/shutdown and 574 VM migration arrival/departure, the NVE can make optimizations if it 575 is notified of such information. For example, if NVE knows the 576 incoming activate message caused by migration rather than VM 577 creation, some mechanisms may be employed or triggered to make sure 578 the dynamic configurations or provisionings on the destination NVE 579 same as those on the source NVE for the migrated VM, for example 580 multicast group memberships. 582 4. Hypervisor-to-NVE Signaling Protocol requirements 584 Req-1: The protocol is able to run between the hypervisor and its 585 associated external NVE which may directly connected or bridged in 586 split-NVE architecture. 588 Req-2: The protocol MUST support the hypervisor initiating a request 589 to its associated external NVE to be connected to a given VN. 591 Req-3: In response to the connection request to a given VN received 592 on NVE's port p as per Req-1, the protocol SHOULD support NVE 593 replying a locally significant tag assigned, for example 802.1Q tag 594 value, to each of the VN it is member of. NVE should keep the record 595 of VN ID, local tag assigned and port p triplet. 597 Req-4: The protocol MUST support the hypervisor initiating a request 598 to associate/disassociate, activate/deactive or clear a TSI instance 599 to a VN on an NVE port. All requests should be logically consistent 600 with text in section 5.2 & 5.3. 602 Req-5: The protocol MUST support the hypervisor initiating a request 603 to add, remove or update addresses associated with a TSI instance 604 which has associated or activated on the external NVE. Addresses can 605 be expressed in different formats, for example, MAC, IP or pair of IP 606 and MAC. 608 Req-6: When any request of the protocol fails, a reason code MUST be 609 provided in the reply. 611 Req-7: The protocol MAY support the hypervisor explicitly informing 612 NVE when a migration starts. It may help NVE to differentiate a new 613 associated/activated TSI resulting from VM creation or VM migration. 615 Req-8: The protocol SHOULD be extensible to carry more parameters to 616 meet future requirements, for example, QoS settings. 618 There are multiple candidate protocols probably with some simple 619 extensions that can be used to exchange signaling information between 620 hypervisor and external NVE. They include VDP [IEEE 802.1Qbg], LLDP, 621 XMPP, and HTTP REST. Multiple factors influence the choice of 622 protocol(s), for example, connection between hypervisor and external 623 NVE is L2 or L3. Appendix A illustrates VDP for reader's information. 625 5. Security Considerations 627 NVEs must ensure that only properly authorized Tenant Systems are 628 allowed to join and become a part of any specific Virtual Network. In 629 addition, NVEs will need appropriate mechanisms to ensure that any 630 hypervisor wishing to use the services of an NVE are properly 631 authorized to do so. One design point is whether the hypervisor 632 should supply the NVE with necessary information (e.g., VM addresses, 633 VN information, or other parameters) that the NVE uses directly, or 634 whether the hypervisor should only supply a VN ID and an identifier 635 for the associated VM (e.g., its MAC address), with the NVE using 636 that information to obtain the information needed to validate the 637 hypervisor-provided parameters or obtain related parameters in a 638 secure manner. 640 6. IANA Considerations 642 No IANA action is required. RFC Editor: please delete this section 643 before publication. 645 7. Acknowledgements 647 This document was initiated and merged from the drafts draft-kreeger- 648 nvo3-hypervisor-nve-cp, draft-gu-nvo3-tes-nve-mechanism and draft- 649 kompella-nvo3-server2nve. Thanks to all the co-authors and 650 contributing members of those drafts. 652 8. References 654 8.1 Normative References 656 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 657 Requirement Levels", BCP 14, RFC 2119, March 1997. 659 8.2 Informative References 661 [I-D.ietf-nvo3-overlay-problem-statement] Narten, T., Gray, E., 662 Black, D., Fang, L., Kreeger, L., and M. Napierala, 663 "Problem Statement: Overlays for Network Virtualization", 664 draft-ietf-nvo3-overlay-problem-statement-04 (work in 665 progress), July 2013. 667 [I-D.ietf-nvo3-framework] Lasserre, M., Balus, F., Morin, T., Bitar, 668 N., and Y. Rekhter, "Framework for DC Network 669 Virtualization", draft-ietf-nvo3-framework-05 (work in 670 progress), January 2014. 672 [I-D.ietf-nvo3-nve-nva-cp-req] Kreeger, L., Dutt, D., Narten, T., and 673 D. Black, "Network Virtualization NVE to NVA Control 674 Protocol Requirements", draft-ietf-nvo3-nve-nva-cp-req-01 675 (work in progress), October 2013. 677 [I-D.ietf-nvo3-arch] Black, D., Narten, T., et al, "An Architecture 678 for Overlay Networks (NVO3)", draft-narten-nvo3-arch, work 679 in progress. 681 [I-D.ietf-opsawg-vmm-mib] Asai H., MacFaden M., Schoenwaelder J., 682 Shima K., Tsou T., "Management Information Base for 683 Virtual Machines Controlled by a Hypervisor", draft-ietf- 684 opsawg-vmm-mib-00 (work in progress), February 2014. 686 [IEEE 802.1Qbg] IEEE, "Media Access Control (MAC) Bridges and Virtual 687 Bridged Local Area Networks - Amendment 21: Edge Virtual 688 Bridging", IEEE Std 802.1Qbg, 2012 690 [8021Q] IEEE, "Media Access Control (MAC) Bridges and Virtual Bridged 691 Local Area Networks", IEEE Std 802.1Q-2011, August, 2011 693 Appendix A. IEEE 802.1Qbg VDP Illustration 695 VDP has the format shown in Figure A.1. Virtual Station Interface (VSI) 696 is an interface to a virtual station that is attached to a downlink port 697 of an internal bridging function in server. VSI's VDP packet will be 698 handled by an external bridge. VDP is the controlling protocol running 699 between the hypervisor and the external bridge. 701 +--------+--------+------+----+----+------+------+------+-----------+ 702 |TLV type|TLV info|Status|VSI |VSI |VSIID | VSIID|Filter|Filter Info| 703 | 7b |str len | |Type|Type|Format| | Info | | 704 | | 9b | 1oct |ID |Ver | | |format| | 705 | | | |3oct|1oct| 1oct |16oct |1oct | M oct | 706 +--------+--------+------+----+----+------+------+------+-----------+ 707 | | | | | 708 | | |<--VSI type&instance-->|<----Filter------>| 709 | | |<------------VSI attributes-------------->| 710 |<--TLV header--->|<-------TLV info string = 23 + M octets--------->| 712 Figure A.1: VDP TLV definitions 714 There are basically four TLV types. 716 1. Pre-Associate: Pre-Associate is used to pre-associate a VSI instance 717 with a bridge port. The bridge validates the request and returns a 718 failure Status in case of errors. Successful pre-association does not 719 imply that the indicated VSI Type or provisioning will be applied to any 720 traffic flowing through the VSI. The pre-associate enables faster 721 response to an associate, by allowing the bridge to obtain the VSI Type 722 prior to an association. 724 2. Pre-Associate with resource reservation: Pre-Associate with Resource 725 Reservation involves the same steps as Pre-Associate, but on successful 726 pre-association also reserves resources in the Bridge to prepare for a 727 subsequent Associate request. 729 3. Associate: The Associate creates and activates an association between 730 a VSI instance and a bridge port. The Bridge allocates any required 731 bridge resources for the referenced VSI. The Bridge activates the 732 configuration for the VSI Type ID. This association is then applied to 733 the traffic flow to/from the VSI instance. 735 4. Deassociate: The de-associate is used to remove an association 736 between a VSI instance and a bridge port. Pre-Associated and Associated 737 VSIs can be de-associated. De-associate releases any resources that were 738 reserved as a result of prior Associate or Pre-Associate operations for 739 that VSI instance. 741 Deassociate can be initiated by either side and the rest types of 742 messages can only be initiated by the server side. 744 Some important flag values in VDP Status field: 746 1. M-bit (Bit 5): Indicates that the user of the VSI (e.g., the VM) is 747 migrating (M-bit = 1) or provides no guidance on the migration of the 748 user of the VSI (M-bit = 0). The M-bit is used as an indicator relative 749 to the VSI that the user is migrating to. 751 2. S-bit (Bit 6): Indicates that the VSI user (e.g., the VM) is 752 suspended (S-bit = 1) or provides no guidance as to whether the user of 753 the VSI is suspended (S-bit = 0). A keep-alive Associate request with 754 S-bit = 1 can be sent when the VSI user is suspended. The S-bit is used 755 as an indicator relative to the VSI that the user is migrating from. 757 The filter information format currently supports 4 types as the 758 following. 760 1. VID Filter Info format 761 +---------+------+-------+--------+ 762 | #of | PS | PCP | VID | 763 |entries |(1bit)|(3bits)|(12bits)| 764 |(2octets)| | | | 765 +---------+------+-------+--------+ 766 |<--Repeated per entry->| 768 Figure A.2 VID Filter Info format 770 2. MAC/VID filter format 771 +---------+--------------+------+-------+--------+ 772 | #of | MAC address | PS | PCP | VID | 773 |entries | (6 octets) |(1bit)|(3bits)|(12bits)| 774 |(2octets)| | | | | 775 +---------+--------------+------+-------+--------+ 776 |<--------Repeated per entry---------->| 778 Figure A.3 MAC/VID filter format 780 3. GroupID/VID filter format 781 +---------+--------------+------+-------+--------+ 782 | #of | GroupID | PS | PCP | VID | 783 |entries | (4 octets) |(1bit)|(3bits)|(12bits)| 784 |(2octets)| | | | | 785 +---------+--------------+------+-------+--------+ 786 |<--------Repeated per entry---------->| 788 Figure A.4 GroupID/VID filter format 790 4. GroupID/MAC/VID filter format 791 +---------+----------+-------------+------+-----+--------+ 792 | #of | GroupID | MAC address | PS | PCP | VID | 793 |entries |(4 octets)| (6 octets) |(1bit)|(3b )|(12bits)| 794 |(2octets)| | | | | | 795 +---------+----------+-------------+------+-----+--------+ 796 |<-------------Repeated per entry------------->| 797 Figure A.5 GroupID/MAC/VID filter format 799 The null VID can be used in the VDP Request sent from the hypervisor to 800 the external bridge. Use of the null VID indicates that the set of VID 801 values associated with the VSI is expected to be supplied by the Bridge. 802 The Bridge can obtain VID values from the VSI Type whose identity is 803 specified by the VSI Type information in the VDP Request. The set of VID 804 values is returned to the station via the VDP Response. The returned VID 805 value can be a locally significant value. When GroupID is used, it is 806 equivalent to the VN ID in NVO3. GroupID will be provided by the 807 hypervisor to the bridge. The bridge will map GroupID to a locally 808 significant VLAN ID. 810 The VSIID in VDP request that identify a VM can be one of the following 811 format: IPV4 address, IPV6 address, MAC address, UUID or locally 812 defined. 814 We compare VDP against the requirements in the following Figure A.6. It 815 should be noted that the comparison is conceptual. Detail parameters 816 checking is not performed. 818 +------+-----------+----------------------------------------------+ 819 | Req | VDP | remarks | 820 | | supported?| | 821 +------+-----------+----------------------------------------------+ 822 | Req-1| partial |support directly connected but not bridged | 823 +------+-----------+----------------------------------------------+ 824 | Req-2| Yes |VN is represented by GroupID | 825 +------+-----------+----------------------------------------------+ 826 | Req-3| Yes |VID=NULL in request and bridge returns the | 827 | | |assigned value in response | 828 +------+-----------+------------------------+---------------------+ 829 | | | requiments | VDP equivalence | 830 | | +------------------------+---------------------+ 831 | Req-4| partial | associate/disassociate| pre-asso/de-asso | 832 | | | activate/deactivate | associate/nil | 833 | | | clear | de-associate | 834 +------+-----------+------------------------+---------------------+ 835 | Req-5| partial | VDP can handle MAC addresses properly. For IP| 836 | | | addresses, it is not clearly specified. | 837 +------+-----------+----------------------------------------------+ 838 | | | | 839 | Req-6| Yes | Error type indicated in Status in response | 840 +------+-----------+----------------------------------------------+ 841 | Req-7| Yes | M bit indicated in Status in request | 842 +------+-----------+----------------------------------------------+ 843 | | | For certain information,e.g. new filter info | 844 | Req-8| partial | format, VDP can easily be extended. For some,| 845 | | | extensibility may be limited. | 846 +------+-----------+----------------------------------------------+ 848 Figure A.6 Compare VDP with the requirements 850 Authors' Addresses 852 Yizhou Li 853 Huawei Technologies 854 101 Software Avenue, 855 Nanjing 210012 856 China 858 Phone: +86-25-56625409 859 EMail: liyizhou@huawei.com 861 Lucy Yong 862 Huawei Technologies, USA 864 Email: lucy.yong@huawei.com 865 Lawrence Kreeger 866 Cisco 868 Email: kreeger@cisco.com 870 Thomas Narten 871 IBM 873 Email: narten@us.ibm.com 874 David Black 875 EMC 877 Email: david.black@emc.com