idnits 2.17.1 draft-ellard-nfsv4-federated-fs-03.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** It looks like you're using RFC 3978 boilerplate. You should update this to the boilerplate described in the IETF Trust License Policy document (see https://trustee.ietf.org/license-info), which is required now. -- Found old boilerplate from RFC 3978, Section 5.1 on line 18. -- Found old boilerplate from RFC 3978, Section 5.5, updated by RFC 4748 on line 1092. -- Found old boilerplate from RFC 3979, Section 5, paragraph 1 on line 1103. -- Found old boilerplate from RFC 3979, Section 5, paragraph 2 on line 1110. -- Found old boilerplate from RFC 3979, Section 5, paragraph 3 on line 1116. 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 : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust Copyright Line does not match the current year -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (August 5, 2008) is 5743 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) ** Obsolete normative reference: RFC 3530 (Obsoleted by RFC 7530) ** Obsolete normative reference: RFC 4346 (Obsoleted by RFC 5246) Summary: 4 errors (**), 0 flaws (~~), 1 warning (==), 7 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group D. Ellard 3 Internet-Draft C. Everhart 4 Intended status: Standards Track NetApp, Inc. 5 Expires: February 6, 2009 R. Tewari 6 M. Naik 7 IBM Almaden 8 August 5, 2008 10 Requirements for Federated File Systems 11 draft-ellard-nfsv4-federated-fs-03.txt 13 Status of this Memo 15 By submitting this Internet-Draft, each author represents that any 16 applicable patent or other IPR claims of which he or she is aware 17 have been or will be disclosed, and any of which he or she becomes 18 aware will be disclosed, in accordance with Section 6 of BCP 79. 20 Internet-Drafts are working documents of the Internet Engineering 21 Task Force (IETF), its areas, and its working groups. Note that 22 other groups may also distribute working documents as Internet- 23 Drafts. 25 Internet-Drafts are draft documents valid for a maximum of six months 26 and may be updated, replaced, or obsoleted by other documents at any 27 time. It is inappropriate to use Internet-Drafts as reference 28 material or to cite them other than as "work in progress." 30 The list of current Internet-Drafts can be accessed at 31 http://www.ietf.org/ietf/1id-abstracts.txt. 33 The list of Internet-Draft Shadow Directories can be accessed at 34 http://www.ietf.org/shadow.html. 36 This Internet-Draft will expire on February 6, 2009. 38 Copyright Notice 40 Copyright (C) The IETF Trust (2008). 42 Abstract 44 This draft describes and lists the functional requirements of a 45 federated file system and defines related terms. Our intent is to 46 use this draft as a starting point and refine it, with input and 47 feedback from the file system community and other interested parties, 48 until we reach general agreement. We will then begin, again with the 49 help of any interested parties, to define standard, open federated 50 file system protocols that satisfy these requirements and are 51 suitable for implementation and deployment. 53 Table of Contents 55 1. Requirements notation . . . . . . . . . . . . . . . . . . . . 3 56 2. Draft Goals . . . . . . . . . . . . . . . . . . . . . . . . . 4 57 3. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 58 4. Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 59 5. Examples and Discussion . . . . . . . . . . . . . . . . . . . 8 60 5.1. Create a Fileset and its FSL(s) . . . . . . . . . . . . . 8 61 5.1.1. Creating a Fileset and a FSN . . . . . . . . . . . . . 8 62 5.1.2. Adding a Replica of a Fileset . . . . . . . . . . . . 9 63 5.2. Junction Resolution . . . . . . . . . . . . . . . . . . . 9 64 5.3. Junction Creation . . . . . . . . . . . . . . . . . . . . 11 65 6. Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 66 7. Proposed Requirements . . . . . . . . . . . . . . . . . . . . 16 67 7.1. Basic Assumptions . . . . . . . . . . . . . . . . . . . . 16 68 7.2. Requirements . . . . . . . . . . . . . . . . . . . . . . . 19 69 8. Non-Requirements . . . . . . . . . . . . . . . . . . . . . . . 25 70 9. IANA Requirements . . . . . . . . . . . . . . . . . . . . . . 26 71 10. Security Considerations . . . . . . . . . . . . . . . . . . . 27 72 11. References . . . . . . . . . . . . . . . . . . . . . . . . . . 28 73 11.1. Normative References . . . . . . . . . . . . . . . . . . . 28 74 11.2. Informational References . . . . . . . . . . . . . . . . . 28 75 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 29 76 Intellectual Property and Copyright Statements . . . . . . . . . . 30 78 1. Requirements notation 80 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 81 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 82 document are to be interpreted as described in [RFC2119]. 84 Note, that this is a requirements document, and in many instances 85 where these words are used in this document they refer to qualities 86 of a specification for a system that satisfies the document, or 87 requirements of a system that matches that specification. These 88 cases are distinguished when there is potential for ambiguity. 90 2. Draft Goals 92 This draft describes and lists the functional requirements of a 93 federated file system and defines related terms. Our intent is to 94 use this draft as a starting point and refine it, with input and 95 feedback from the file system community and other interested parties, 96 until we reach general agreement. We will then begin, again with the 97 help of any interested parties, to define standard, open federated 98 file system protocols that satisfy these requirements and are 99 suitable for implementation and deployment. 101 We do not describe the mechanisms that might be used to implement 102 this functionality except in cases where specific mechanisms, in our 103 opinion, follow inevitably from the requirements. Our focus is on 104 the interfaces between the entities of the system, not on the 105 protocols or their implementations. 107 For the first version of this document, we are focused on the 108 following questions: 110 o Are any "MUST" requirements missing? 112 o Are there any "MUST" requirements that should be "SHOULD" or 113 "MAY"? 115 o Are there any "SHOULD" requirements that should be "MAY"? 117 o Are there better ways to articulate the requirements? 119 3. Overview 121 Today, there are collections of fileservers that inter-operate to 122 provide a single namespace comprised of filesystem resources provided 123 by different members of the collection, joined together with inter- 124 filesystem junctions. The namespace can either be assembled at the 125 fileservers, the clients, or by an external namespace service -- the 126 mechanisms used to assemble the namespace may vary depending on the 127 filesystem access protocol used by the client. 129 These fileserver collections are, in general, administered by a 130 single administrative entity. This administrator builds the 131 namespace out of the filesystem resources and junctions. There are 132 also singleton servers that export some or all of their filesystem 133 resources, but which do not contain junctions to other filesystems. 135 Current server collections that provide a shared namespace usually do 136 so by means of a service that maps filesystem names to filesystem 137 locations. We refer to this as a namespace database service (NSDB). 138 In some distributed file systems, this service is embodied as a 139 volume location database (VLDB), and may be implemented by LDAP, NIS, 140 or any number of other mechanisms. 142 We use the term "fileset" to represent the abstraction of a 143 filesystem. The fileset abstraction implies very little about how 144 the fileset is implemented, although in the simplest case a fileset 145 can be implemented by an exported filesystem. A fileset is a 146 directory tree that may contain files and references, called 147 "junction", to other filesets. Each fileset has a fileset globally 148 unique name (FSN) that is used as an identifier for the fileset. 149 Each implementation of a given fileset is specified by its fileset 150 location (FSL). 152 The primary purpose of the NSDB service is to provide a level of 153 indirection between the FSN the FSLs. If the NSDB service permits 154 updates to the set of mappings, then the FSLs may be changed (e.g., 155 moved or replicated) in a manner that is transparent to the referring 156 fileset and its server(s). 158 Current approaches are unsuitable to build common namespaces across 159 systems with multiple administrative domains and multiple NSDB nodes. 160 An approach which requires changing existing NSDB nodes to 161 collaborate or replacing them with a single NSDB node, while 162 possible, is not desirable. 164 Figure Figure 1 shows an example of a federation. This federation 165 has two members, named ALPHA and BETA. Federation members may 166 contain an arbitrary number of file servers and NSDB nodes; in this 167 illustration ALPHA and BETA each have three servers and one NSDB 168 node. 170 +----------------------+ +----------------------+ 171 | Federation Member | | Federation Member | 172 | ALPHA | | BETA | 173 | | | | 174 | | | | 175 | +------------+ | | +------------+ | 176 | | NSDB | | | | NSDB | | 177 | | | | | | | | 178 | +------------+ | | +------------+ | 179 | | | | 180 | | | | 181 | | | | 182 | +----------+ | | +----------+ | 183 | | | | | | | | 184 | +-- | Servers | | | +-- | Servers | | 185 | | | | | | | | | | 186 | +-- | | | | | +-- | | | | 187 | | | +----------+ | | | | +----------+ | 188 | | | | | | | | | | 189 | | +----------+ | | | +----------+ | 190 | | | | | | | | 191 | +----------+ | | +----------+ | 192 +----------------------+ +----------------------+ 194 A federation with two members, ALPHA and BETA. ALPHA and BETA each 195 have their own NSDB node and several file servers, but are 196 administered separately. 198 Figure 1 200 4. Purpose 202 Our objective is to specify a set of interfaces (and corresponding 203 protocols) by which such fileservers and collections of fileservers, 204 with different administrators, can form a federation of fileservers 205 and NSDB nodes that provides a namespace composed of the filesets 206 hosted on the different fileservers and fileserver collections. 208 It should be possible, using a system that implements the interfaces, 209 to share a common namespace across all the fileservers in the 210 federation. It should also be possible for different fileservers in 211 the federation to project different namespaces and enable clients to 212 traverse them. 214 Such a federation may contain an arbitrary number of NSDB nodes, each 215 belonging to a different administrative entity, and each providing 216 the mappings that define a part of a namespace. Such a federation 217 may also have an arbitrary number of administrative entities, each 218 responsible for administering a subset of the servers and NSDB nodes. 219 Acting in concert, the administrators should be able to build and 220 administer this multi-fileserver, multi-collection namespace. 222 Each singleton server can be presumed to provide its own NSDB node, 223 for example with a trivial mapping to local FSLs. 225 It is not the intent of the federation to guarantee namespace 226 consistency across all client views. Since different parts of the 227 namespace may be administered by different entities, it is possible 228 that a client could be accessing a stale area of the namespace 229 managed by one entity because a part of the namespace above it, 230 managed by another entity, has changed. 232 5. Examples and Discussion 234 In this section we provide examples and discussion of the basic 235 operations facilitated by the federated file system protocol: 236 creating a fileset, adding a replica of a fileset, resolving a 237 junction, and creating a junction. 239 5.1. Create a Fileset and its FSL(s) 241 A fileset is the abstraction of a set of files and their containing 242 directory tree. The fileset abstraction is the fundamental unit of 243 data management in the federation. This abstraction is implemented 244 by an actual directory tree whose root location is specified by a 245 fileset location (FSL). 247 In this section, we describe the basic requirements for starting with 248 a directory tree and creating a fileset that can be used in the 249 federation protocols. Note that we do not assume that the process of 250 creating a fileset requires any transformation of the files or the 251 directory hierarchy. The only thing that is required by this process 252 is assigning the fileset a fileset name (FSN) and expressing the 253 location(s) of the implementation of the fileset as FSL(s). 255 There are many possible variations to this procedure, depending on 256 how the FSN that binds the FSL is created, and whether other replicas 257 of the fileset exist, are known to the federation, and need to be 258 bound to the same FSN. 260 It is easiest to describe this in terms of how to create the initial 261 implementation of the fileset, and then describe how to add replicas. 263 5.1.1. Creating a Fileset and a FSN 265 1. Choose the NSDB node that will keep track of the FSL(s) and 266 related information for the fileset. 268 2. Request that the NSDB node register a new FSN for the fileset. 270 The FSN may either be chosen by the NSDB node or by the server. 271 The latter case is used if the fileset is being restored, perhaps 272 as part of disaster recovery, and the server wishes to specify 273 the FSN in order to permit existing junctions that reference that 274 FSN to work again. 276 At this point, the FSN exists, but its location is unspecified. 278 3. Send the FSN, the local volume path, the export path, and the 279 export options for the local implementation of the fileset to the 280 NSDB node. Annotations about the FSN or the location may also be 281 sent. 283 The NSDB node records this info and creates the initial FSL for 284 the fileset. 286 5.1.2. Adding a Replica of a Fileset 288 Adding a replica is straightforward: the NSDB node and the FSN are 289 already known. The only remaining step is to add another FSL. 291 Note that the federation interfaces do not include methods for 292 creating or managing replicas: this is assumed to be a platform- 293 dependent operation (at least at this time). The only interface 294 required is the ability to register or remove the registration of 295 replicas for a fileset. 297 5.2. Junction Resolution 299 A fileset may contain references to other filesets. These references 300 are represented by junctions. If a client requests access to a 301 fileset object that is a junction, the server resolves the junction 302 to discover the FSL(s) that implements the referenced fileset. 304 There are many possible variations to this procedure, depending on 305 how the junctions are represented and how the information necessary 306 to perform resolution is represented by the server. In this example, 307 we assume that the only thing directly expressed by the junction is 308 the junction key; its mapping to FSN can be kept local to the server 309 hosting the junction. 311 Step 5 is the only step that interacts directly with the federation 312 interfaces. The rest of the steps may use platform-specific 313 interfaces. 315 1. The server determines that the object being accessed is a 316 junction. 318 2. The server determines the junction key for the junction. 320 3. Using the junction key, the server does a local lookup to find 321 the FSN of the target fileset. 323 4. Using the FSN, the server finds the NSDB node responsible for the 324 target object. 326 5. The server contacts that NSDB node and asks for the set of FSLs 327 that implement the target FSN. The NSDB node responds with a set 328 of FSLs. 330 6. The server converts the FSL to the location type used by the 331 client (e.g., fs_location for NFSv4, as described in [RFC3530]). 333 7. The server redirects (in whatever manner is appropriate for the 334 client) the client to the location(s). 336 These steps are illustrated in Figure 2. The client sends request 1 337 to server X, in federation member ALPHA, in an attempt to reference 338 an object (which appears to the client as a directory). Server X 339 recognizes that the referenced object is actually a junction that 340 refers to a directory in a different fileset. Server X finds, from 341 the FSN in the junction, that the NSDB responsible for knowing the 342 location of the target of the junction is the NSDB of federation 343 member BETA. Server X sends request 2 to the NSDB of BETA, asking 344 for the current location of the directory. The NSDB sends response 3 345 to server X, telling the server that the directory is located on 346 server Y. Server X sends response 4 to the client, indicating that 347 the directory is in a "new" location on server Y. The client then 348 sends request 5 to server Y, repeating the initial request. 350 Given the current requirements and definitions, this resolution 351 method MUST work. However, there is no requirement that this is the 352 only resolution method that can be used. This method may be used as 353 the fallback when all else fails (or, for a simple implementation, it 354 could be the only method). This is a degenerate implementation of 355 the NSDB service as a simple composition of NSDB nodes; we expect 356 that large federations will use more sophisticated methods to share 357 the FSN and FSL information among multiple NSDB nodes. 359 +---------------+ 360 | | 361 | Client | >--------------------------+ 362 | | | 363 +---------------+ | 364 v ^ | 365 +-----+---+-------------+ +-----------------+-----+ 366 | | | Federation| |Federation | | 367 | | | member | |member | | 368 | | | ALPHA | |BETA | | 369 | | | | | | | 370 | | | | | | | 371 | | | | | | | 372 | | | | | | | 373 | | | | | +---------+ | | 374 | | | +---------+------+-> | | | | 375 | | | | | | | NSDB Y | | | 376 | | | | +-----+------+-< | | | | 377 | | | | | | | +---------+ | | 378 | | | | | | | | | 379 | | | | | | | | | 380 | | | | | | | | | 381 | 1| 4| 2| 3| | | 5| | 382 | v ^ ^ v | | v | 383 | +---------------+ | | +---------------+ | 384 | | | | | | | | 385 | | Server X | | | | Server Y | | 386 | | | | | | | | 387 | +---------------+ | | +---------------+ | 388 | | | | 389 +-----------------------+ +-----------------------+ 391 Figure 2 393 5.3. Junction Creation 395 Given a local path, a remote export and a path relative to that 396 export, create a junction from the local path to the path within the 397 remote export. 399 There are many possible variations to this procedure, depending on 400 how the junctions are represented and how the information necessary 401 to perform resolution is represented by the server. In this example, 402 we assume that the only thing directly expressed by the junction is 403 the junction key; its mapping to FSN can be kept local to the server 404 hosting the junction. 406 Step 1 is the only step that uses the federation interfaces. The 407 rest of the steps may use platform-specific interfaces. 409 1. Contact the server named by the export and ask for the FSN for 410 the fileset, given its path relative to that export. 412 2. Create a new local junction key. 414 3. Insert, in the local junction info table, a mapping from the 415 local junction key to the FSN. 417 4. Insert the junction, at the given path, into the local 418 filesystem. 420 6. Glossary 422 The phrase "USING THE FEDERATION INTERFACES" implies that the 423 subsequent requirement must be satisfied, in its entirety, via the 424 federation interfaces. 426 Administrator: user with the necessary authority to initiate 427 administrative tasks on one or more servers. 429 Admin entity: A server or agent that administers a collection of 430 fileservers and persistently stores the namespace information. 432 Client: Any client that accesses the fileserver data using a 433 supported filesystem access protocol. 435 Federation: A set of server collections and singleton servers that 436 use a common set of interfaces and protocols in order to provide 437 to their clients a federated namespace accessible through a 438 filesystem access protocol. 440 Fileserver: A server exporting a filesystem via a network filesystem 441 access protocol. 443 Fileset: The abstraction of a set of files and their containing 444 directory tree. A fileset is the fundamental unit of data 445 management in the federation. 447 Note that all files within a fileset are descendants of one 448 directory, and that filesets do not span filesystems. 450 Filesystem: A self-contained unit of export for a fileserver, and 451 the mechanism used to implement filesets. The fileset does not 452 need to be rooted at the root of the filesystem, nor at the export 453 point for the filesystem. 455 A single filesystem MAY implement more than one fileset, if the 456 client protocol and the fileserver permit this. 458 Filesystem access protocol: A network filesystem access protocol 459 such as NFSv2 [RFC1094], NFSv3 [RFC1813], NFSv4 [RFC3530], or 460 CIFS. 462 FSL (Fileset location): The location of the implementation of a 463 fileset at a particular moment in time. A FSL MUST be something 464 that can be translated into a protocol-specific description of a 465 resource that a client can access directly, such as a fs_location 466 (for NFSv4), or share name (for CIFS). Note that not all FSLs 467 need to be explicitly exported as long as they are contained 468 within an exported path on the fileserver. 470 FSN (Fileset name): A platform-independent and globally unique name 471 for a fileset. Two FSLs that implement replicas of the same 472 fileset MUST have the same FSN, and if a fileset is migrated from 473 one location to another, the FSN of that fileset MUST remain the 474 same. 476 Junction: A filesystem object used to link a directory name in the 477 current fileset with an object within another fileset. The 478 server-side "link" from a leaf node in one fileset to the root of 479 another fileset. 481 Junction key: The UUID of a fileset, used as a key to lookup an FSN 482 within an NSDB node or a local table of information about 483 junctions. 485 Namespace: A filename/directory tree that a sufficiently-authorized 486 client can observe. 488 NSDB (Namespace Database Service): A service that maps FSNs to FSLs. 489 The NSDB may also be used to store other information, such as 490 annotations for these mappings and their components. 492 NSDB Node: The name or location of a server that implements part of 493 the NSDB service and is responsible for keeping track of the FSLs 494 (and related info) that implement a given partition of the FSNs. 496 Referral: A server response to a client access that directs the 497 client to evaluate the current object as a reference to an object 498 at a different location (specified by an FSL) in another fileset, 499 and possibly hosted on another fileserver. The client re-attempts 500 the access to the object at the new location. 502 Replica: A replica is a redundant implementation of a fileset. Each 503 replica shares the same FSN, but has a different FSL. 505 Replicas may be used to increase availability or performance. 506 Updates to replicas of the same fileset MUST appear to occur in 507 the same order, and therefore each replica is self-consistent at 508 any moment. 510 We do not assume that updates to each replica occur simultaneously 511 If a replica is offline or unreachable, the other replicas may be 512 updated. 514 Server Collection: A set of fileservers administered as a unit. A 515 server collection may be administered with vendor-specific 516 software. 518 The namespace provided by a server collection could be part of the 519 federated namespace. 521 Singleton Server: A server collection containing only one server; a 522 stand-alone fileserver. 524 7. Proposed Requirements 526 Note that the requirements are described in terms of correct behavior 527 by all entities. We do not address the requirements of the system in 528 the presence of faults. 530 7.1. Basic Assumptions 532 Several of the requirements are so fundamental that we treat them as 533 basic assumptions; if any of these assumptions are violated, the rest 534 of the requirements must be reviewed in their entirety. 536 A1: The federation protocols do not require any changes to existing 537 client-facing protocols, and MAY be extended to incorporate new 538 client-facing protocols. 540 A2: A client SHOULD NOT require any a priori knowledge of the 541 general structure or composition of the federation. 543 The client may require some specific knowledge in order to find 544 and access an instance of the fileset that defines the root of 545 its view of the namespace. As the client traverses the 546 namespace, the client discovers the information it needs in 547 order to locate the filesets it accesses. 549 A3: All requirements MUST be satisfiable via the federation 550 protocols and the standard protocols used by the fileservers 551 (i.e., NFS, CIFS, DNS, etc). 553 USING THE FEDERATION INTERFACES, a federation operation that 554 requires an interaction between two (or more) entities that are 555 members of the federation MUST be possible without requiring any 556 proprietary protocols. 558 A4: All the entities participating in a federation operation MUST be 559 able to authenticate each other. 561 All principals (clients, users, administrator of a singleton or 562 server collection, hosts, NSDB nodes, etc) that can assume a 563 role defined by the federation protocol can identify themselves 564 to each other via an authentication mechanism. This mechanism 565 is not defined or further described in this document. 567 The authority of a principal to request that a second principal 568 perform a specific operation is ultimately determined by the 569 second. Authorization may be partitioned by server collection 570 or set of servers as well as by operation. For example, if a 571 user has administrative privileges on one server in the 572 federation, this does not imply that they have administrative 573 privileges (or, for that matter, any privileges whatsoever) on 574 any other server in the federation. 576 In order to access the functionality provided by the federation 577 interfaces, it may be necessary to have elevated privileges or 578 authorization. The authority required by different operations 579 may be different. For example, the authority required to query 580 the NSDB about the FSLs bound to an FSN may be different than 581 the authority required to change the bindings of that FSN. 583 An operation attempted by an unauthorized entity MUST fail in a 584 manner that indicates that the failure was due to insufficient 585 authorization. 587 This document does not enumerate the authorization necessary for 588 any operation. 590 A5: The federation protocols MUST NOT require changes to existing 591 authentication/authorization mechanisms in use at the 592 fileservers for client-facing protocols. 594 A user's view of the namespace may be limited by the 595 authentication and authorization privileges it has on the 596 different fileservers in the federation. As such, users may 597 only be able to traverse the parts of the namespace that they 598 have access to. 600 The federation protocols do not impose any restrictions on how 601 users are represented within the federation. For example, a 602 single enterprise could employ a common identity for users 603 across the federation. A grid environment could utilize user 604 mapping or translations across different administrative domains. 606 A6: In a federated system, we assume that a FSN MUST express, or can 607 be used to discover, the following two pieces of information: 609 1. The location of the NSDB node that is responsible for 610 knowing the filesystem location(s) (FSLs) of the named 611 fileset. 613 The NSDB node must be specified because there may be many 614 NSDB nodes in a federation. We do not assume that any 615 single entity knows the location of all of the NSDB nodes, 616 and therefore exhaustive search is not an option. 618 There are several ways in which a fileserver can locate the 619 NSDB node responsible for a given fileset. One approach, 620 given a DNS infrastructure, is to specify the location of 621 the NSDB node by the FQDN of the server hosting the NSDB 622 node. Another approach is to use a separate DNS-style 623 hierarchy to resolve the location of the NSDB node. 625 2. The junction key. 627 The junction key is the index used by the NSDB node to 628 identify the FSN of the target fileset. 630 There are several ways to represent junction keys. One 631 approach could use 128-bit UUIDs as described described in 632 [RFC4122]. 634 As an example, an FSN could be represented by a URL of the form 635 nsdb.example.com/UUID where nsdb.example.com is the FQDN of the 636 server hosting the NSDB node and UUID is the string 637 representation of the junction key. 639 Note that it is not assumed that it is always required for a 640 server to contact the NSDB node specified by the FSN in order to 641 find the FSLs. The relevant information stored in that NSDB 642 node may also be cached local to the server or on a proxy NSDB 643 node "near" the server. 645 A7: All federation servers and NSDB nodes are assumed to execute the 646 federation protocols correctly. The behavior of the federation 647 is undefined in the case of Byzantine behavior by any federation 648 server or NSDB node. 650 A8: The locations of federation services (such as NSDBs and FSLs) 651 can be specified in a manner such that they can be correctly 652 interpreted by all members of the federation that will access 653 them. 655 For example, if an NSDB node is specified by a FQDN, then this 656 implies that every member of the federation that needs to access 657 this NSDB node can resolve this FQDN to an IP address for that 658 NSDB node. (It is not necessary that the FQDN always resolve to 659 the same address; the same service may appear at different 660 addresses on different networks.) 662 It is the responsibility of each federation member to ensure 663 that the resources it wishes to expose have accessible network 664 locations and that the necessary resolution mechanisms (i.e., 665 DNS) are given the necessary data to perform the resolution 666 correctly. 668 7.2. Requirements 670 R1: Requirements of each FSN: 672 a. Each FSN MUST be globally unique. 674 b. The FSN MUST be sufficiently descriptive to locate an 675 instance of the fileset it names within the federation at 676 any time. 678 c. An FSN is a name of a fileset. (An FSL is not the name of 679 a fileset, but only a locator of an instance of a fileset 680 at some point in time. For example, the same FSL may 681 implement different filesets at different times.) 683 + If a fileset instance is moved to a new location, it 684 will have a new FSL, but its FSN is unchanged. 686 + An instance of a different fileset may be placed at a 687 FSL previously occupied by an instance of a different 688 fileset. 690 d. If a fileset instance is migrated to another location, the 691 FSN remains the same in the new location. 693 e. If the fileset is replicated using the federation 694 interfaces, then all of the replicas have the same FSN. 696 Not all filesets in the federation are required to have a FSN 697 or be reachable by a FSL. Only those filesets that are the 698 target of a junction (as described in R3) are required to have 699 an FSN. 701 NOTE: this requirement has been called into question. 703 R2: USING THE FEDERATION INTERFACES, it MUST be possible to create 704 an FSN for a fileset, and it must be possible to bind an FSL to 705 that FSN. These operations are NSDB operations and do not 706 require any action on the part of an NFS server. 708 It is possible to create an FSN for a fileset that has not 709 actually been created. It is also possible to bind a 710 nonexistant FSL to an FSN. It is also possible to create a 711 fileset without assigning it an FSN. The binding between an 712 FSN and an FSL is defined entirely within the context of the 713 NSDB; the servers do not "know" whether the filesets they host 714 have been assigned FSNs (or, if so, what those FSNs are). 716 The requirement that filesets can exist prior to being assigned 717 an FSN, and the requirement that FSNs can exist independent of 718 filesets are intended to simplify the construction of the 719 namespace in a convenient manner. For example, they permit an 720 admin to assign FSNs to existing filesets and thereby 721 incorporate existing filesets into the namespace. They also 722 permit the structure of the namespace to be defined prior to 723 creation of the component filesets. In either case, it is the 724 responsibility of the entity updating the NSDB with FSNs and 725 FSN-to-FSL mappings to ensure that the namespace is constructed 726 in a consistent manner. (The simplest way to accomplish this 727 is to ensure that the FSN and FSN-to-FSL mappings are always 728 recorded in the NSDB prior to the creation of any junctions 729 that refer to that FSN.) 731 a. An administrator MAY specify the entire FSN (including both 732 the NSDB node location and the junction key) of the newly- 733 created FSL, or the administrator MAY specify only the NSDB 734 node and have the system choose the junction key. 736 The admin can choose to specify the FSN explicitly in order 737 to recreate a lost fileset with a given FSN (for example, 738 as part of disaster recovery). It is an error to assign an 739 FSN that is already in use by an active fileset. 741 Note that creating a replica of an existing filesystem is 742 NOT accomplished by assigning the FSN of the filesystem you 743 wish to replicate to a new filesystem. 745 b. USING THE FEDERATION INTERFACES, it MUST be possible to 746 create a federation FSL by specifying a specific local 747 volume, path, export path, and export options. 749 R3: USING THE FEDERATION INTERFACES, and given the FSN of a target 750 fileset, it MUST be possible to create a junction to that 751 fileset at a named place in another fileset. 753 After a junction has been created, clients that access the 754 junction transparently interpret it as a reference to the 755 FSL(s) that implement the FSN associated with the junction. 757 a. It SHOULD be possible to have more than one junction whose 758 target is a given fileset. In other words, it SHOULD be 759 possible to mount a fileset at multiple named places. 761 b. If the fileset in which the junction is created is 762 replicated, then the junction MUST eventually appear in all 763 of its replicas. 765 The operation of creating a junction within a fileset is 766 treated as an update to the fileset, and therefore obey the 767 general rules about updates to replicated filesets. 769 R4: USING THE FEDERATION INTERFACES, it MUST be possible to delete 770 a specific junction from a fileset. 772 If a junction is deleted, clients who are already viewing the 773 fileset referred to by the junction after traversing the 774 junction MAY continue to view the old namespace. They might 775 not discover that the junction no longer exists (or has been 776 deleted and replaced with a new junction, possibly referring to 777 a different FSN). 779 After a junction is deleted, another object with the same name 780 (another junction, or an ordinary filesystem object) may be 781 created. 783 The operation of deleting a junction within a fileset is 784 treated as an update to the fileset, and therefore obey the 785 general rules about updates to replicated filesets. 787 R5: USING THE FEDERATION INTERFACES, it MUST be possible to 788 invalidate an FSN. 790 a. If a junction refers to an FSN that is invalid, attempting 791 to traverse the junction MUST fail. 793 An FSN that has been invalidated MAY become valid again if the 794 FSN is recreated (i.e., as part of a disaster recovery 795 process). 797 If an FSN is invalidated, clients who are already viewing the 798 fileset named by the FSN MAY continue to view the old 799 namespace. They might not discover that the FSN is no longer 800 valid until they try to traverse a junction that refers to it. 802 R6: USING THE FEDERATION INTERFACES, it MUST be possible to 803 invalidate a FSL. 805 a. An invalid FSL MUST NOT be returned as the result of 806 resolving a junction. 808 An FSL that has been invalidated MAY become valid again if the 809 FSL is recreated (i.e., as part of a disaster recovery 810 process). 812 If an FSL is invalidated, clients who are already viewing the 813 fileset implemented by the FSL MAY continue to use that FSL. 814 They might not discover that the FSL is no longer valid until 815 they try to traverse a junction that refers to the fileset 816 implemented by the FSL. 818 Note that invalidating an FSL does not imply that the 819 underlying export or share (depending on the file access 820 protocol in use) is changed in any way -- it only changes the 821 mappings from FSNs to FSLs on the NSDB. 823 R7: It MUST be possible for the federation of servers to provide 824 multiple namespaces. 826 R8: USING THE FEDERATION INTERFACES, it MUST be possible to perform 827 queries about the state of objects relevant to the 828 implementation of the federation namespace. 830 It MUST be possible to query the fileserver named in an FSL to 831 discover whether a junction exists at a given path within that 832 FSL. 834 R9: The projected namespace (and the objects named by the 835 namespace) MUST be accessible to clients via at least one 836 standard filesystem access protocol. 838 a. The namespace SHOULD be accessible to clients via the CIFS 839 protocol. 841 b. The namespace SHOULD be accessible to clients via the NFSv4 842 protocol as described in [RFC3530]. 844 c. The namespace SHOULD be accessible to clients via the NFSv3 845 protocol as described in [RFC1813]. 847 d. The namespace SHOULD be accessible to clients via the NFSv2 848 protocol as described in [RFC1094]. 850 It must be understood that some of these protocols, such as 851 NFSv3 and NFSv2, have no innate ability to access a namespace 852 of this kind. Where such protocols have been augmented with 853 other protocols and mechanisms (such as autofs or amd for 854 NFSv3) to provide an extended namespace, we propose that these 855 protocols and mechanisms may be used, or extended, in order to 856 satisfy the requirements given in this draft, and different 857 clients may use different mechanisms. 859 R10: USING THE FEDERATION INTERFACES, it MUST be possible to modify 860 the NSDB mapping from an FSN to a set of FSLs to reflect the 861 migration from one FSL to another. 863 R11: FSL migration SHOULD have little or no impact on the clients, 864 but this is not guaranteed across all federation members. 866 Whether FSL migration is performed transparently depends on 867 whether the source and destination servers are able to do so. 868 It is the responsibility of the administrator to recognize 869 whether or not the migration will be transparent, and advise 870 the system accordingly. The federation, in turn, MUST advise 871 the servers to notify their clients, if necessary. 873 For example, on some systems, it may be possible to migrate a 874 fileset from one system to another with minimal client impact 875 because all client-visible metadata (inode numbers, etc) are 876 preserved during migration. On other systems, migration might 877 be quite disruptive. 879 R12: USING THE FEDERATION INTERFACES, it MUST be possible to modify 880 the NSDB mapping from an FSN to a set of FSLs to reflect the 881 addition/removal of a replica at a given FSL. 883 R13: Replication SHOULD have little or no negative impact on the 884 clients. 886 Whether FSL replication is performed transparently depends on 887 whether the source and destination servers are able to do so. 888 It is the responsibility of the administrator initiating the 889 replication to recognize whether or not the replication will be 890 transparent, and advise the federation accordingly. The 891 federation MUST advise the servers to notify their clients, if 892 necessary. 894 For example, on some systems, it may be possible to mount any 895 FSL of an FSN read/write, while on other systems, there may be 896 any number of read-only replicas but only one FSL that can be 897 mounted read-write. 899 R14: USING THE FEDERATION INTERFACES, it SHOULD be possible to 900 annotate the objects and relations managed by the federation 901 protocol with arbitrary name/value pairs. 903 These annotations are not used by the federation protocols -- 904 they are intended for use by higher-level protocols. For 905 example, an annotation that might be useful for a system 906 administrator browsing the federation would be the "owner" of 907 each FSN (i.e., "this FSN is for the home directory of Joe 908 Smith."). As another example, the annotations may express 909 hints used by the clients (such as priority information for 910 NFSv4.1). 912 Both FSNs and FSLs may be annotated. For example, an FSN 913 property might be "This is Joe Smith's home directory", and an 914 FSL property might be "This instance of the FSN is at the 915 remote backup site." 917 a. USING THE FEDERATION INTERFACES, it MUST be possible to 918 query the system to find the annotations for a junction. 920 b. USING THE FEDERATION INTERFACES, it MUST be possible to 921 query the system to find the annotations for a FSN. 923 c. USING THE FEDERATION INTERFACES, it MUST be possible to 924 query the system to find the annotations for a FSL. 926 8. Non-Requirements 928 N1: It is not necessary for the namespace to be known by any 929 specific fileserver. 931 In the same manner that clients do not need to have a priori 932 knowledge of the structure of the namespace or its mapping onto 933 federation members, the projected namespace can exist without 934 individual fileservers knowing the entire organizational 935 structure, or, indeed, without knowing exactly where in the 936 projected namespace the filesets they host exist. 938 Fileservers do need to be able to handle referrals from other 939 fileservers, but they do not need to know what path the client 940 was accessing when the referral was generated. 942 N2: It is not necessary for updates and accesses to the federation 943 data to occur in transaction or transaction-like contexts. 945 One possible requirement that is omitted from our current list 946 is that updates and accesses to the data stored in the NSDB (or 947 individual NSDB nodes) occur within a transaction context. We 948 were not able to agree whether the benefits of transactions are 949 worth the complexity they add (both to the specification and its 950 eventual implementation) but this topic is open for discussion. 952 Below is the the draft of a proposed requirement that provides 953 transactional semantics: 955 "There MUST be a way to ensure that sequences of operations, 956 including observations of the namespace (including finding 957 the locations corresponding to a set of FSNs) and changes to 958 the namespace or related data stored in the system (including 959 the creation, renaming, or deletion of junctions, and the 960 creation, altering, or deletion of mappings between FSN and 961 filesystem locations), can be performed in a manner that 962 provides predictable semantics for the relationship between 963 the observed values and the effect of the changes." 965 "It MUST be possible to protect sequences of operations by 966 transactions with NSDB-wide or server-wide ACID semantics." 968 9. IANA Requirements 970 This document has no actions for IANA. 972 10. Security Considerations 974 Assuming the Internet threat model, the federated resolution 975 mechanism described in this document MUST be implemented in such a 976 way to prevent loss of CONFIDENTIALITY, DATA INTEGRITY and PEER 977 ENTITY AUTHENTICATION, as described in [RFC3552]. 979 CONFIDENTIALITY may be violated if an unauthorized party is able to 980 eavesdrop on the communication between authorized servers and NSDB 981 nodes and thereby learn the locations or other information about FSNs 982 that they would not be authorized to discover via direct queries. 983 DATA INTEGRITY may be compromised if a third party is able to 984 undetectably alter the contents of the communication between servers 985 and NSDB nodes. PEER ENTITY AUTHENTICATION is defeated if one server 986 can masquerade as another server without proper authority, or if an 987 arbitrary host can masquerade as a NSDB node. 989 Well-established techniques for providing authenticated channels may 990 be used to defeat these attacks, and the protocol MUST support at 991 least one of them. 993 For example, if LDAP is used to implement the query mechanism 994 [RFC4511], then TLS may be used to provide both authentication and 995 integrity [RFC4346] [RFC4513]. If the query protocol is implemented 996 on top of ONC/RPC, then RPCSEC_GSS may be used to fill the same role 997 [RFC2203] [RFC2743]. 999 11. References 1001 11.1. Normative References 1003 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1004 Requirement Levels", BCP 14, RFC 2119, March 1997. 1006 [RFC2203] Eisler, M., Chiu, A., and L. Ling, "RPCSEC_GSS Protocol 1007 Specification", RFC 2203, September 1997. 1009 [RFC2743] Linn, J., "Generic Security Service Application Program 1010 Interface Version 2, Update 1", RFC 2743, January 2000. 1012 [RFC3530] Shepler, S., Callaghan, B., Robinson, D., Thurlow, R., 1013 Beame, C., Eisler, M., and D. Noveck, "Network File System 1014 (NFS) version 4 Protocol", RFC 3530, April 2003. 1016 [RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC 1017 Text on Security Considerations", BCP 72, RFC 3552, 1018 July 2003. 1020 [RFC4122] Leach, P., Mealling, M., and R. Salz, "A Universally 1021 Unique IDentifier (UUID) URN Namespace", RFC 4122, 1022 July 2005. 1024 [RFC4346] Dierks, T. and E. Rescorla, "The Transport Layer Security 1025 (TLS) Protocol Version 1.1", RFC 4346, April 2006. 1027 [RFC4511] Sermersheim, J., "Lightweight Directory Access Protocol 1028 (LDAP): The Protocol", RFC 4511, June 2006. 1030 [RFC4513] Harrison, R., "Lightweight Directory Access Protocol 1031 (LDAP): Authentication Methods and Security Mechanisms", 1032 RFC 4513, June 2006. 1034 11.2. Informational References 1036 [RFC1094] Nowicki, B., "NFS: Network File System Protocol 1037 specification", RFC 1094, March 1989. 1039 [RFC1813] Callaghan, B., Pawlowski, B., and P. Staubach, "NFS 1040 Version 3 Protocol Specification", RFC 1813, June 1995. 1042 Authors' Addresses 1044 Daniel Ellard 1045 NetApp, Inc. 1046 1601 Trapelo Rd, Suite 16 1047 Waltham, MA 02451 1048 US 1050 Phone: +1 781-768-5421 1051 Email: ellard@netapp.com 1053 Craig Everhart 1054 NetApp, Inc. 1055 7301 Kit Creek Rd 1056 Research Triangle Park, NC 27709 1057 US 1059 Phone: +1 919-476-5320 1060 Email: everhart@netapp.com 1062 Renu Tewari 1063 IBM Almaden 1064 650 Harry Rd 1065 San Jose, CA 95120 1066 US 1068 Email: tewarir@us.ibm.com 1070 Manoj Naik 1071 IBM Almaden 1072 650 Harry Rd 1073 San Jose, CA 95120 1074 US 1076 Email: manoj@almaden.ibm.com 1078 Full Copyright Statement 1080 Copyright (C) The IETF Trust (2008). 1082 This document is subject to the rights, licenses and restrictions 1083 contained in BCP 78, and except as set forth therein, the authors 1084 retain all their rights. 1086 This document and the information contained herein are provided on an 1087 "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS 1088 OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND 1089 THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS 1090 OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF 1091 THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED 1092 WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. 1094 Intellectual Property 1096 The IETF takes no position regarding the validity or scope of any 1097 Intellectual Property Rights or other rights that might be claimed to 1098 pertain to the implementation or use of the technology described in 1099 this document or the extent to which any license under such rights 1100 might or might not be available; nor does it represent that it has 1101 made any independent effort to identify any such rights. Information 1102 on the procedures with respect to rights in RFC documents can be 1103 found in BCP 78 and BCP 79. 1105 Copies of IPR disclosures made to the IETF Secretariat and any 1106 assurances of licenses to be made available, or the result of an 1107 attempt made to obtain a general license or permission for the use of 1108 such proprietary rights by implementers or users of this 1109 specification can be obtained from the IETF on-line IPR repository at 1110 http://www.ietf.org/ipr. 1112 The IETF invites any interested party to bring to its attention any 1113 copyrights, patents or patent applications, or other proprietary 1114 rights that may cover technology that may be required to implement 1115 this standard. Please address the information to the IETF at 1116 ietf-ipr@ietf.org. 1118 Acknowledgment 1120 Funding for the RFC Editor function is provided by the IETF 1121 Administrative Support Activity (IASA).