idnits 2.17.1 draft-hellwig-nfsv4-scsi-layout-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (March 23, 2015) is 3322 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Possible downref: Non-RFC (?) normative reference: ref. 'LEGAL' ** Obsolete normative reference: RFC 5661 (Obsoleted by RFC 8881) -- Possible downref: Non-RFC (?) normative reference: ref. 'SAM-4' -- Possible downref: Non-RFC (?) normative reference: ref. 'SPC3' Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 NFSv4 C. Hellwig 3 Internet-Draft March 23, 2015 4 Intended status: Standards Track 5 Expires: September 24, 2015 7 Parallel NFS (pNFS) SCSI Layout 8 draft-hellwig-nfsv4-scsi-layout-01.txt 10 Abstract 12 Parallel NFS (pNFS) extends Network File Sharing version 4 (RFC5661) 13 to allow clients to directly access file data on the storage used by 14 the NFSv4 server. This ability to bypass the server for data access 15 can increase both performance and parallelism, but requires 16 additional client functionality for data access, some of which is 17 dependent on the class of storage used. The main pNFS operations 18 document specifies storage-class-independent extensions to NFS, the 19 pNFS Block/Volume Layout (RFC5663) specifies the additional 20 extensions for use of pNFS with block-and volume-based storage, while 21 this document provides extensions to the pNFS Block/Volume Layout 22 document to provide reliable fencing and better device 23 discoverability for SCSI based shared storage devices. 25 Status of this Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at http://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on September 24, 2015. 42 Copyright Notice 44 Copyright (c) 2015 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 Table of Contents 59 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.1. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . 3 61 1.2. Conventions Used in This Document . . . . . . . . . . . . 3 62 1.3. Code Components Licensing Notice . . . . . . . . . . . . . 3 63 1.4. XDR Description . . . . . . . . . . . . . . . . . . . . . 3 64 2. SCSI Layout Description . . . . . . . . . . . . . . . . . . . 5 65 2.1. GETDEVICEINFO . . . . . . . . . . . . . . . . . . . . . . 6 66 2.1.1. Model . . . . . . . . . . . . . . . . . . . . . . . . 6 67 2.1.2. Volume Identification . . . . . . . . . . . . . . . . 7 68 2.1.3. Volume Topology . . . . . . . . . . . . . . . . . . . 7 69 2.2. Client Fencing . . . . . . . . . . . . . . . . . . . . . . 9 70 2.2.1. PRs - Key Generation . . . . . . . . . . . . . . . . . 9 71 2.2.2. PRs - MDS Registration and Reservation . . . . . . . . 9 72 2.2.3. PRs - Client Registration . . . . . . . . . . . . . . 9 73 2.2.4. PRs - Fencing Action . . . . . . . . . . . . . . . . . 10 74 2.2.5. Client Recovery After a Fence Action . . . . . . . . . 10 75 3. Security Considerations . . . . . . . . . . . . . . . . . . . 10 76 4. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 11 77 5. Normative References . . . . . . . . . . . . . . . . . . . . . 11 78 Appendix A. Acknowledgments . . . . . . . . . . . . . . . . . . . 11 79 Appendix B. RFC Editor Notes . . . . . . . . . . . . . . . . . . 12 80 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 12 82 1. Introduction 84 In the parallel Network File System (pNFS), the metadata server 85 returns Layout Type structures that describe where file data is 86 located. There are different Layout Types for different storage 87 systems and methods of arranging data on storage devices. This 88 document extends the pNFS Block/Volume Layout [RFC5663] with a closer 89 integration into the the SCSI Architecture Model ([SAM-4]) to provide 90 a generic fencing method and more scalable device discovery. 92 1.1. Scope 94 This document only specifies an updated version of the layout- 95 specific GETDEVICEINFO XDR response, and a new mandatory fencing 96 method for SCSI devices, but refers to [RFC5663] for the basic 97 principle of operation, as well as the layout specific XDR data 98 structures for the LAYOUTGET and LAYOUTCOMMIT operations. This 99 document does not directly interact with [RFC6688], although the 100 mechanisms described in this document also achieve the goals of 101 [RFC6688], and do so in a more robust fashion that does not depend on 102 the cooperation of the systems involved. Thus, the mechanisms 103 specified in [RFC6688] are not necessary for a pNFS SCSI layout type 104 implementation. 106 1.2. Conventions Used in This Document 108 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 109 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 110 document are to be interpreted as described in [RFC2119]. 112 1.3. Code Components Licensing Notice 114 The external data representation (XDR) description and scripts for 115 extracting the XDR description are Code Components as described in 116 Section 4 of "Legal Provisions Relating to IETF Documents" [LEGAL]. 117 These Code Components are licensed according to the terms of Section 118 4 of "Legal Provisions Relating to IETF Documents". 120 1.4. XDR Description 122 This document contains the XDR [RFC4506] description of the NFSv4.1 123 SCSI layout protocol. The XDR description is embedded in this 124 document in a way that makes it simple for the reader to extract into 125 a ready-to-compile form. The reader can feed this document into the 126 following shell script to produce the machine readable XDR 127 description of the NFSv4.1 SCSI layout: 129 #!/bin/sh 130 grep '^ *///' $* | sed 's?^ */// ??' | sed 's?^ *///$??' 132 That is, if the above script is stored in a file called "extract.sh", 133 and this document is in a file called "spec.txt", then the reader can 134 do: 136 sh extract.sh < spec.txt > flex_files_prot.x 138 The effect of the script is to remove leading white space from each 139 line, plus a sentinel sequence of "///". 141 The embedded XDR file header follows. Subsequent XDR descriptions, 142 with the sentinel sequence are embedded throughout the document. 144 Note that the XDR code contained in this document depends on types 145 from the NFSv4.1 nfs4_prot.x file [RFC5662]. This includes both nfs 146 types that end with a 4, such as offset4, length4, etc., as well as 147 more generic types such as uint32_t and uint64_t. 149 /// /* 150 /// * This code was derived from draft-hellwig-nfsv4-scsi-layout 151 /// * Please reproduce this note if possible. 152 /// */ 153 /// /* 154 /// * Copyright (c) 2015 IETF Trust and the persons identified 155 /// * as the document authors. All rights reserved. 156 /// * 157 /// * Redistribution and use in source and binary forms, with 158 /// * or without modification, are permitted provided that the 159 /// * following conditions are met: 160 /// * 161 /// * - Redistributions of source code must retain the above 162 /// * copyright notice, this list of conditions and the 163 /// * following disclaimer. 164 /// * 165 /// * - Redistributions in binary form must reproduce the above 166 /// * copyright notice, this list of conditions and the 167 /// * following disclaimer in the documentation and/or other 168 /// * materials provided with the distribution. 169 /// * 170 /// * - Neither the name of Internet Society, IETF or IETF 171 /// * Trust, nor the names of specific contributors, may be 172 /// * used to endorse or promote products derived from this 173 /// * software without specific prior written permission. 174 /// * 175 /// * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 176 /// * AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED 177 /// * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 178 /// * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS 179 /// * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO 180 /// * EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE 181 /// * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 182 /// * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 183 /// * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR 184 /// * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 185 /// * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF 186 /// * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, 187 /// * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 188 /// * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF 189 /// * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 190 /// */ 191 /// 192 /// /* 193 /// * nfs4_scsi_layout_prot.x 194 /// */ 195 /// 196 /// %#include "nfs4_block_layout_prot.x" 197 /// %#include "nfsv41.h" 198 /// 200 2. SCSI Layout Description 202 The layout4 type defined in [RFC5662] is extended with a new value as 203 follows: 205 enum layouttype4 { 206 LAYOUT4_NFSV4_1_FILES = 1, 207 LAYOUT4_OSD2_OBJECTS = 2, 208 LAYOUT4_BLOCK_VOLUME = 3, 209 LAYOUT4_SCSI = 0x80000005 210 [[RFC Editor: please modify the LAYOUT4_SCSI 211 to be the layouttype assigned by IANA]] 212 }; 214 struct layout_content4 { 215 layouttype4 loc_type; 216 opaque loc_body<>; 217 }; 219 struct layout4 { 220 offset4 lo_offset; 221 length4 lo_length; 222 layoutiomode4 lo_iomode; 223 layout_content4 lo_content; 224 }; 226 This document defines structure associated with the layouttype4 value 227 LAYOUT4_SCSI. [RFC5661] specifies the loc_body structure as an XDR 228 type "opaque". The opaque layout is uninterpreted by the generic 229 pNFS client layers, but obviously must be interpreted by the Layout 230 Type implementation. All structures behind this opaque value are 231 identical to those defined in [RFC5663]. 233 2.1. GETDEVICEINFO 235 /// /* 236 /// * Code sets from SPC-3. 237 /// */ 238 /// enum pnfs_scsi_code_set { 239 /// PS_CODE_SET_BINARY = 1, 240 /// PS_CODE_SET_ASCII = 2, 241 /// PS_CODE_SET_UTF8 = 3 242 /// }; 243 /// 244 /// /* 245 /// * Designator types from taken from SPC-3. 246 /// * 247 /// * Other values are allocated in SPC-3, but not mandatory to 248 /// * implement or aren't logical unit names. 249 /// */ 250 /// enum pnfs_scsi_designator_type { 251 /// PS_DESIGNATOR_EUI64 = 2, 252 /// PS_DESIGNATOR_NAA = 3, 253 /// PS_DESIGNATOR_NAME = 8 254 /// }; 255 /// 256 /// /* 257 /// * Logical unit name + reservation key. 258 /// */ 259 /// struct pnfs_scsi_base_volume_info4 { 260 /// pnfs_scsi_code_set sbv_code_set; 261 /// pnfs_scsi_designator_type sbv_designator_type; 262 /// opaque sbv_designator<>; 263 /// uint32_t sbv_pr_key; 264 /// }; 265 /// 267 2.1.1. Model 269 GETDEVICEINFO calls are handled exactly the same way as specified in 270 [RFC5663]. The "pnfs_scsi_volume_type4" data structure returned by 271 the server as the storage-protocol-specific opaque field da_addr_body 272 in the "device_addr4" structure by a successful GETDEVICEINFO 273 operation [RFC5661] is a strict superset of the 274 "pnfs_block_volume_type" structured defined by [RFC5663]. 276 2.1.2. Volume Identification 278 SCSI targets implementing [SPC3] export unique logical unit names for 279 each logical unit through the Device Identification VPD page which 280 can be obtained using the INQUIRY command. This document uses a 281 subset of this information to identify logical units backing pNFS 282 SCSI layouts. It is similar to the "Identification Descriptor Target 283 Descriptor" specified in [SPC3], but limits the allowed values to 284 those that uniquely identify a logical unit. Device Identification 285 VPD page descriptors used to identify logical units for use with pNFS 286 SCSI layouts must adhere to the following restrictions: 288 1. The "ASSOCIATION" must be set to 0 (The DESIGNATOR field is 289 associated with the addressed logical unit). 291 2. The "DESIGNATOR TYPE" must be set to one of three values 292 explicitly listed in the "pnfs_scsi_designator_type" 293 enumerations. 295 The "CODE SET" VPD page field is stored in the "sbv_code_set" field 296 of the "pnfs_scsi_base_volume_info4" structure, the "DESIGNATOR TYPE" 297 is stored in "sbv_designator_type", and the DESIGNATOR is stored in 298 "sbv_designator". Due to the use of a XDR array the "DESIGNATOR 299 LENGTH" field does not need to be set separately. Only certain 300 combinations of "sbv_code_set" and "sbv_designator_type" are valid, 301 please refer to [SPC3] for details, and note that ASCII may be used 302 as the code set for UTF-8 text that contains only ASCII characters. 303 Note that a Device Identification VPD page MAY contain multiple 304 descriptors with the same association, code set and designator type. 305 NFS clients thus MUST iterate the descriptors until a match for 306 "sbv_code_set", "sbv_designator_type" and "sbv_designator" is found, 307 or until the end of VPD page. 309 Additionally the server returns a Persistent Reservation key in the 310 "sbv_pr_key" field. See Section 2.2 for more details on the use of 311 Persistent Reservations. 313 2.1.3. Volume Topology 315 The pNFS SCSI server volume topology is expressed as an arbitrary 316 combination of base volume types enumerated in the following data 317 structures. The individual components of the topology are contained 318 in an array and components may refer to other components by using 319 array indices. 321 /// enum pnfs_scsi_volume_type4 { 322 /// PNFS_SCSI_VOLUME_SIMPLE = 323 /// PNFS_BLOCK_VOLUME_SIMPLE , /* invalid */ 324 /// PNFS_SCSI_VOLUME_SLICE = /* see RFC5663 */ 325 /// PNFS_BLOCK_VOLUME_SLICE, 326 /// PNFS_SCSI_VOLUME_CONCAT = /* see RFC5663 */ 327 /// PNFS_BLOCK_VOLUME_CONCAT, 328 /// PNFS_SCSI_VOLUME_STRIPE = /* see RFC5663 */ 329 /// PNFS_BLOCK_VOLUME_STRIPE, 330 /// PNFS_SCSI_VOLUME_BASE = 4 /* SCSI LU */ 331 /// }; 332 /// 334 /// 335 /// union pnfs_scsi_volume4 switch (pnfs_scsi_volume_type4 type) { 336 /// case PNFS_SCSI_VOLUME_SIMPLE: 337 /// pnfs_block_simple_volume_info4 sv_simple_info; 338 /// case PNFS_SCSI_VOLUME_SLICE: 339 /// pnfs_block_slice_volume_info4 sv_slice_info; 340 /// case PNFS_SCSI_VOLUME_CONCAT: 341 /// pnfs_block_concat_volume_info4 sv_concat_info; 342 /// case PNFS_SCSI_VOLUME_STRIPE: 343 /// pnfs_block_stripe_volume_info4 sv_stripe_info; 344 /// case PNFS_SCSI_VOLUME_BASE: 345 /// pnfs_scsi_base_volume_info4 sv_base_info; 346 /// }; 347 /// 349 /// /* scsi layout specific type for da_addr_body */ 350 /// struct pnfs_scsi_deviceaddr4 { 351 /// pnfs_scsi_volume4 sda_volumes<>; /* array of volumes */ 352 /// }; 353 /// 355 All rules for ordering and formation of a "pnfs_scsi_deviceaddr4" 356 structure are identical to those for a "pnfs_block_deviceaddr4" 357 structure in [RFC5663], except that the new 358 pnfs_scsi_base_volume_info4 PNFS_SCSI_VOLUME_BASE case is used in 359 place of the pnfs_block_simple_volume_info4 PNFS_BLOCK_VOLUME_SIMPLE 360 case as the base structure. A PNFS_BLOCK_VOLUME_SIMPLE element MUST 361 NOT be referenced by a pnfs_scsi_deviceaddr4, but is preserved for 362 XDR level compatibility. 364 2.2. Client Fencing 366 [RFC5663] suggests using either LUN masking or cooperative clients to 367 implement client fencing. The first implementation requires the 368 server and the storage device to have a common way to address a 369 client, which is impossible when the NFS and storage connection don't 370 share a network, and requires a non-standardized control protocol 371 between the MDS and the storage device. The second implementation 372 relies on a cooperative client, which is not robust. 374 Instead this document specifies a new SCSI-specific fencing protocol 375 using Persistent Reservations (PRs), similar to the fencing method 376 used by existing shared disk file systems. By placing a PR of type 377 "Exclusive Access - All Registrants" on each SCSI logical unit 378 exported to pNFS clients the MDS prevents access from any client that 379 does not have an outstanding device device ID that gives the client a 380 reservation key to access the logical unit, and allows the MDS to 381 revoke access to the logic unit at any time. 383 2.2.1. PRs - Key Generation 385 To allow fencing individual systems, each system must use a unique 386 Persistent Reservation key. [SPC3] does not specify a way to 387 generate keys. This document assigns the burden to generate unique 388 keys to the MDS, which must generate a key for itself before 389 exporting a volume, and one for each client that accesses a volume. 390 The MDS MAY either generate a key for each client that accesses logic 391 units exported by the MDS, or generate a key for each [logical unit, 392 client] combination. If using a single key per client, the MDS needs 393 to be aware of the per-client fencing granularity. 395 2.2.2. PRs - MDS Registration and Reservation 397 Before returning a PNFS_SCSI_VOLUME_BASE volume to the client, the 398 MDS needs to prepare the volume for fencing using PRs. This is done 399 by registering the reservation generated for the MDS with the device 400 using the "PERSISTENT RESERVE OUT" command with a service action of 401 "REGISTER", followed by a "PERSISTENT RESERVE OUT" command, with a 402 service action of "RESERVE" and the type field set to 8h (Exclusive 403 Access - All Registrants). To make sure all I_T nexuses are 404 registered, the MDS SHOULD set the "All Target Ports" (ALL_TG_PT) bit 405 when registering the key, or otherwise ensure the registration is 406 performed for each initiator port. 408 2.2.3. PRs - Client Registration 410 Before performing the first IO to a device returned from a 411 GETDEVICEINFO operation the client will register the registration key 412 returned in sbv_pr_key with the storage device by issuing a 413 "PERSISTENT RESERVE OUT" command with a service action of REGISTER 414 with the "SERVICE ACTION RESERVATION KEY" set to the reservation key 415 returned in sbv_pr_key. To make sure all I_T nexus are registered, 416 the client SHOULD set the "All Target Ports" (ALL_TG_PT) bit when 417 registering the key, or otherwise ensure the registration is 418 performed for each initiator port. 420 When a client stops using a device earlier returned by GETDEVICEINFO 421 it MUST unregister the earlier registered key by issuing a 422 "PERSISTENT RESERVE OUT" command with a service action of "REGISTER" 423 with the "RESERVATION KEY" set to the earlier registered reservation 424 key. 426 2.2.4. PRs - Fencing Action 428 In case of a non-responding client the MDS MUST fence the client by 429 issuing a "PERSISTENT RESERVE OUT" command with the service action 430 set to "PREEMPT" or "PREEMPT AND ABORT", the reservation key field 431 set to the server's reservation key, the service action reservation 432 key field set to the reservation key associated with the non- 433 responding client, and the type field set to 8h (Exclusive Access - 434 All Registrants). 436 After the MDS preempts a client, all client I/O to the logical unit 437 fails. The client should at this point return any layout that refers 438 to the device ID that points to the logical unit. Note that the 439 client can distinguish I/O errors due to fencing from other errors 440 based on the "RESERVATION CONFLICT" status. Refer to [SPC3] for 441 details. 443 2.2.5. Client Recovery After a Fence Action 445 A client that detects I/O errors on the storage devices MUST commit 446 through the MDS, return all outstanding layouts for the device, 447 forget the device ID and unregister the reservation key. Future 448 GETDEVICEINFO calls may refer to the storage device again, in which 449 case a new registration will be performed. 451 3. Security Considerations 453 The security considerations in [RFC5663] apply to this document as 454 well. 456 4. IANA Considerations 458 IANA is requested to assign a new pNFS layout type in the pNFS Layout 459 Types Registry as follows (the value 5 is suggested): Layout Type 460 Name: LAYOUT4_SCSI Value: 0x00000005 RFC: RFCTBD10 How: L (new layout 461 type) Minor Versions: 1 463 5. Normative References 465 [LEGAL] IETF Trust, "Legal Provisions Relating to IETF Documents", 466 November 2008, . 469 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 470 Requirement Levels", March 1997. 472 [RFC4506] Eisler, M., "XDR: External Data Representation Standard", 473 STD 67, RFC 4506, May 2006. 475 [RFC5661] Shepler, S., Ed., Eisler, M., Ed., and D. Noveck, Ed., 476 "Network File System (NFS) Version 4 Minor Version 1 477 Protocol", RFC 5661, January 2010. 479 [RFC5662] Shepler, S., Ed., Eisler, M., Ed., and D. Noveck, Ed., 480 "Network File System (NFS) Version 4 Minor Version 1 481 External Data Representation Standard (XDR) Description", 482 RFC 5662, January 2010. 484 [RFC5663] Black, D., Ed., Fridella, S., Ed., and J. Glasgow, Ed., 485 "Parallel NFS (pNFS) Block/Volume Layout", RFC 5663, 486 January 2010. 488 [RFC6688] Black, D., Ed., Glasgow, J., and S. Faibish, "Parallel NFS 489 (pNFS) Block Disk Protection", RFC 6688, July 2012. 491 [SAM-4] INCITS Technical Committee T10, "SCSI Architecture Model - 492 4 (SAM-4)", ANSI INCITS 447-2008, ISO/IEC 14776-414, 2008. 494 [SPC3] INCITS Technical Committee T10, "SCSI Primary Commands-3", 495 ANSI INCITS 408-2005, ISO/IEC 14776-453, 2005. 497 Appendix A. Acknowledgments 499 David Black, Robert Elliott and Tom Haynes provided a throughout 500 review of early drafts of this document, and their input lead to the 501 current form of the document. 503 Appendix B. RFC Editor Notes 505 [RFC Editor: please remove this section prior to publishing this 506 document as an RFC] 508 [RFC Editor: prior to publishing this document as an RFC, please 509 replace all occurrences of RFCTBD10 with RFCxxxx where xxxx is the 510 RFC number of this document] 512 Author's Address 514 Christoph Hellwig 516 Email: hch@lst.de