Debian Codegraph @

rfc1107.html

raw
1 <pre>Network Working Group K. Sollins
2 Request for Comments: 1107 M.I.T. Laboratory for Computer Science
3 July 1989
4
5
6 <span class="h1">A Plan for Internet Directory Services</span>
7
8
9 Table of Contents
10
11 1. Introduction 1
12 1.1. The Issues 1
13 1.2. Project Summary 3
14 2. Goals and Requirements for a White Pages Service 6
15 3. Pre-existing Services 9
16 4. Proposed Approach 11
17 4.1. Stage 1: The Field Test 12
18 4.2. Stage 2: Implementation 17
19 4.3. Stage 3: Deployment 17
20 5. Conclusion 18
21
22 Status of this Memo
23
24 This memo proposes a program to develop a directory service for the
25 Internet. It reports the results of a meeting held in February 1989,
26 which was convened to review requirements and options for such a
27 service. This proposal is offered for comment, and does not
28 represent a committed research activity of the Internet community.
29 Activity in this area is anticipated, and comments should be provided
30 promptly. Distribution of this memo is unlimited.
31
32 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span>
33
34 <span class="h3"><a class="selflink" id="section-1.1" href="#section-1.1">1.1</a>. The Issues</span>
35
36 As part of the planned growth of the Internet (in particular, in
37 support of the full science research community in the U.S.), an
38 increasing need is anticipated for various sorts of directory
39 services. The increase in the size of the community served by the
40 Internet and the burgeoning demands for electronic mail lead to the
41 need for a service to find people's computer mailboxes and other
42 relevant facts, a so-called "White Pages" service. At the user level
43 to date, there have been no such national or international white
44 pages services in general use. As part of building the National
45 Research Network (NRN), it is important that such a service exist,
46 not only within the NRN community, but also crossing the boundaries
47 from the NRN to the more global network community. This will enhance
48 communication not only among computer scientists, but also among
49
50
51
52 <span class="grey">Sollins [Page 1]</span></pre>
53 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span>
54 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
55
56
57 scientists and engineers in other fields as well. Also important and
58 related is a so-called "Yellow Pages" service, which permits the
59 location of Internet resources based on their attributes.
60
61 A "White Pages" service is one in which one can look up people in
62 order to learn information about them for finding them. In its
63 simplest form, a white pages service provides what the white pages
64 telephone book provides. Based on a name, one can find an address
65 and a telephone number. In a network environment, there may be many
66 other kinds of location information, such as electronic mailbox,
67 electronic calendar, or file server, where one might leave a file for
68 the recipient. In addition, the electronic white pages may support a
69 much more sophisticated set of mechanisms for lookup. One might
70 match on a more complex set of attributes than first and last name.
71 In addition, the searching might span more than one local white pages
72 service. There are a number of naming and directory service
73 specifications and implementations in the field. They have differing
74 functionality and mechanisms to address that functionality.
75
76 Within the the world of networking today, there are a number of
77 partial solutions to the directory service problem. Examples of
78 these are the Internet Domain Naming Service (DNS), Clearinghouse,
79 DECnet Network Architecture Naming Service (DNANS), Profile, and
80 X.500. The Domain Naming Service provides a directory service most
81 commonly used for host naming and mail delivery. Clearinghouse and
82 DNANS are respectively the Xerox and DEC corporate naming services,
83 originally for mail delivery, although having other uses as well, in
84 both cases. Profile is part of the work of Larry Peterson to explore
85 descriptive naming in a non-hierarchical structure.
86
87 There is a CCITT recommendation X.500 (ISO DIS 9594), which defines a
88 general directory service. One of its primary goals is the naming
89 service needed for message handling (X.400). While X.500 is still
90 developing, and would need further evolution to cover all the
91 requirements of a service for the Internet, it will have an important
92 impact on the Internet community. It will form the basis of
93 commercial products, and it will almost certainly be the directory
94 service of many parts of the network world, which implies a need to
95 interoperate at a minimum. There is some concern that despite the
96 fact that X.500 is a recognized standard, there are a number of gaps
97 and limitations of the approach, that in turn will cause it to be
98 inadequate for the needs of the NRN.
99
100 In this context, a meeting was held to review current requirements
101 and solutions for directory services. This RFC reports the results
102 of that meeting, including the possibilities for a program of work in
103 this area.
104
105
106
107
108 <span class="grey">Sollins [Page 2]</span></pre>
109 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span>
110 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
111
112
113 For two days, a group representing academic, commercial, and
114 government interests in directory services discussed both alternative
115 candidates for a white pages service and the issues in building any
116 such service. The meeting was kept small by inviting only a small
117 number of representatives of each perspective. By the conclusion of
118 the second day, a consensus was reached on how one could achieve a
119 white pages service in three years. This is summarized in the next
120 section.
121
122 <span class="h3"><a class="selflink" id="section-1.2" href="#section-1.2">1.2</a>. Project Summary</span>
123
124 The consensus of the meeting can be summarized in the following five
125 points:
126
127 1. The standards and implementations are close enough to being
128 complete that it is reasonable to undertake provision of an NRN
129 "White Pages" service.
130
131 2. Although we are close, an effort is needed to experiment with
132 different levels of service, to flesh out the standards, and to
133 develop code.
134
135 3. An initial evaluation experiment is needed before making final
136 detailed plans for a production version of the service.
137
138 4. With strong funding and encouragement, a production service is
139 possible in three years.
140
141 5. It is important to act now to provide a coherent solution.
142 This means both having an impact on the evolving standards
143 and providing a unified, wide-spread solution before a plethora
144 of differing solutions appear.
145
146 Although it has clearcut drawbacks, X.500 was identified as the most
147 likely candidate directory service. The reasons for this are that it
148 has rich semantics and is becoming the accepted international
149 standard. However, there are problems with its incompleteness and
150 with its strict hierarchy. Therefore, in order to explore these and
151 become convinced of its viability, the consensus at the meeting was
152 to propose field trials, as the project's first stage. The field
153 trials would be limited in the user community, perhaps restricted to
154 computer science departments because of their familiarity with the
155 problems, and would be based on experimental or new software. They
156 would include experiments with at least an X.500 implementation,
157 Profile, and DNANS. Each of these services has strong points that
158 must be considered as part of the evaluation. They are:
159
160
161
162
163
164 <span class="grey">Sollins [Page 3]</span></pre>
165 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span>
166 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
167
168
169 X.500: International standard, hierarchy, search rules and
170 filters for searching attributed based names.
171
172 Profile: Descriptive naming with a richer semantics for
173 describing search criteria, an arbitrary network
174 of servers.
175
176 DNANS: Access control, replication, caching, hierarchy.
177
178 In summary, the plan would fall into three stages as follows:
179
180 - Stage 1: Field Trials.
181
182 There are two aspects to the field trials. The first is to
183 explore several different architectures for a white pages
184 service. To this end, implementations of X.500, Profile, and
185 DNANS should be included. The second aspect of the field
186 trials is to distinguish issues inherent in the X.500
187 specification from artifacts of a particular implementation of
188 it. Therefore, if possible, two implementations of X.500
189 should be included. Only one such implementation, Quipu, was
190 identified as developed enough to be included in a field trial
191 at present, but others are under way, and will follow. This
192 stage must also include a careful and objective review of the
193 field trials.
194
195 - Stage 2: Implementation.
196
197 This stage will include work on both the service and user
198 interfaces. The field trials could result in one of a variety
199 of conclusions about the service. These may range from
200 concluding that one or another of the services suits the needs
201 of the NRN to proposing a compromise position based on a
202 combination of shortcomings of any one service and the features
203 of others to address those shortcomings. Because X.500 will
204 become the standard in other domains, an interface to X.500
205 will be necessary. Since all of these implementations are
206 still under development, in order to provide production quality
207 code, more implementation work will be needed.
208
209 Although some work will have been done on the user interfaces,
210 much more will be needed in this stage to provide a variety of
211 interfaces. Much emphasis should be placed on this in Stage 2.
212
213 - Stage 3: Deployment.
214
215 Deployment of the full white pages service requires information
216 gathering in order to fill the directory service, placement of
217
218
219
220 <span class="grey">Sollins [Page 4]</span></pre>
221 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span>
222 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
223
224
225 servers, distribution of and training for use of client code,
226 placement and management of services, and delegation of
227 authority within the service for authority over the contents.
228 Data collection and some delegation of authority as well as
229 training for users of the client code would begin during the
230 field trial. This stage would begin concurrently with the
231 other two. During the second year, detailed planning for
232 deployment must take place. This stage would conclude in three
233 years, at which time widespread deployment would have occurred.
234
235 In order to undertake this three stage program effectively, the group
236 identified the following major projects:
237
238 - Further implementation of code for the field trials.
239
240 In each case (e.g., Quipu, Profile, and DNANS), programs exist,
241 although modifications are likely to be necessary. For
242 example, each will need to be modified to utilize the common
243 file format into which the input data about users will be
244 gathered.
245
246 - Design, development and evaluation of user interfaces.
247
248 - Design and development of data gathering and management tools.
249
250 - Oversight and evaluation of the field trials.
251
252 Careful thought and planning must go into the field trials, to
253 guarantee that we learn what is needed to make an evaluation
254 and to plan for the white pages service. The evaluation must
255 also produce a document that is both a general specification
256 (assuming no one alternative is chosen wholesale) and profiling
257 information, in order for later interoperability and
258 conformance testing.
259
260 - Detailed planning and later management of deployment.
261
262 This includes delegation of authority over parts of the
263 namespace and arbitrating the shape of the namespace
264 (addressing the questions about who gets what sorts of names).
265 This is in addition to the continued and extended data
266 collection and management, distributing the data, placing the
267 code, documentation and user education.
268
269 - Standards participation is an important part of the program.
270
271 It is critical as X.500 changes during the next 4 year study
272 period that the United States take a strong stand on any
273
274
275
276 <span class="grey">Sollins [Page 5]</span></pre>
277 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span>
278 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
279
280
281 changes we envision. It is encumbant on us to utilize
282 effectively the results of the largest field trials of this
283 work in the international arena. The group agreed that this
284 could take up to one half of one person's time in a year.
285
286 - A task force or working group is necessary to provide a forum
287 for communication and discussion.
288
289 It is important to pursue this path now, both to architect a unified
290 solution before a collection of ad hoc solutions is deployed, and to
291 provide effective input into the X.500 standards work based on the
292 field trials.
293
294 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Goals and Requirements for a White Pages Service</span>
295
296 The requirements of a white pages service are the following:
297
298 - Functionality:
299
300 The simple form of a white pages service is straightforward;
301 one should be able to query the service with the name of a
302 person, and have returned attributes of the person such as
303 network mail address and phone number.
304
305 - Correctness of information:
306
307 The information in a white pages service is useless and
308 untrusted if it is not updated regularly. A white pages
309 service will not be used, if the information it provides is out
310 of date or incorrect. This will require a set of management
311 tools. Data integrity is an especially difficult challenge in
312 this area, in contrast with information that is syntactically
313 correct.
314
315 - Size:
316
317 The science and research community has been estimated at ten
318 million users. The number of organizations in the United
319 States is on the order of ten to one hundred thousand.
320
321 - Usage and query rate:
322
323 In comparison with the typical telephone book pattern of about
324 one lookup a week per person, users of electronic mail in the
325 science and research community will send more electronic mail
326 messages than they currently make phone calls, leading to an
327 estimate of ten searches a week per user for electronic as well
328 as paper mail and telephone information. This leads to a query
329
330
331
332 <span class="grey">Sollins [Page 6]</span></pre>
333 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span>
334 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
335
336
337 rate of 10**8 queries per week or 170 per second on average,
338 with much higher peak rates. The average could probably be
339 handled by a single server, but not the peak rates and this
340 would leave little room for growth. Therefore, a distributed,
341 multiple server solution is the only one that make sense.
342
343 - Response time:
344
345 The issue of overall query behavior must be considered
346 carefully. The issue arises when queries, in particular
347 searches, are not limited to tightly constrained sets of
348 entries. Since the number of queries generated will be
349 proportional to the number of users (and the size of the
350 system), the white pages design must avoid costs per query that
351 are related to the size of the system. The consequence,
352 otherwise, will be quadratic behavior in response time.
353
354 The response time of the service must also reflect the expected
355 usage. A phone book style query must respond in the waiting
356 time tolerable to a user, perhaps ten seconds maximum, or one
357 second desirable. If the service is incorporated as a
358 component of a larger service, then the needs of that service
359 determine the response time.
360
361 - Partitioned Authority:
362
363 The white pages service under discussion would be used by a
364 wide variety of organizations, ranging from small and large
365 companies, to network service providers, to government
366 agencies. Many of these would find it unacceptable to delegate
367 the authority over their namespaces to some other organization.
368 Therefore, partitioned authority including some access control,
369 name assignment, and information management must be possible.
370
371 - Access Control:
372
373 The access control required by the white pages falls into two
374 categories, read access control, and write or modify access
375 control. There are at least two reasons that read access
376 control must be available. One is that organizations may
377 require limiting the access to the actual entries or parts of
378 them. This would be comparable to organizations not being
379 willing to distribute their corporate phone books or personnel
380 records. The other reason is that some organizations do not
381 want to publicize or make public their organizational
382 structure. Write and modify access control is necessary
383 because both individuals and organizations may want to prevent
384 inadvertent or malicious creation or modification of
385
386
387
388 <span class="grey">Sollins [Page 7]</span></pre>
389 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span>
390 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
391
392
393 information. Access control is an issue for both organizations
394 wanting to retain local control of personnel information and
395 individuals wanting to control access to private information
396 about themselves.
397
398 - Multiple Transport Protocol Support:
399
400 Within the next three years, one cannot expect all the
401 organizations in the USA to convert to the OSI protocols. On
402 the other hand, some will. It is therefore important that any
403 white pages service provide interfaces on top of both OSI
404 protocols and TCP/IP. There currently exists a partial OSI
405 suite know as ISODE on top of TCP. This is being distributed
406 widely enough that perhaps this should also be supported.
407
408 In addition to these requirements, there are a number of features
409 that would make a white pages service more useful. These are:
410
411 - Additional Functionality:
412
413 Descriptive naming with sophisticated searching based on
414 attributes would support a more flexible human interface than
415 simple name translation. Descriptive naming also would support
416 a general yellow pages style service.
417
418 The form of a yellow pages service is less certain. One
419 definition of a yellow pages service is a directory that stores
420 a number of pre-computed inversions of the directory database,
421 so that entries can be retrieved very efficiently using these
422 predetermined attributes of the data. Another definition of a
423 yellow pages service is one that provides a very powerful set
424 of search primitives, somewhat in common with a relational
425 query language, to support retrieval of entries that match
426 complex attribute conditions. In other words, one view of a
427 yellow pages service is that it is constructed to avoid
428 expensive searches, the other is that it is to facilitate
429 general searches.
430
431 - Accountability:
432
433 Accountability is important both for allocation and recovery of
434 costs. Vendors may provide commercial directory services,
435 therefore depending on accounting as part of their successful
436 commercial ventures.
437
438 - Multiple Interfaces:
439
440 There should be both human and programming interfaces to the
441
442
443
444 <span class="grey">Sollins [Page 8]</span></pre>
445 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span>
446 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
447
448
449 white pages. For example, in addition to human lookups, mail
450 services could effectively use a naming service allow users to
451 include human oriented names than the current electronic mail
452 addresses that are required, such as full domain names.
453
454 - Multiple Clients:
455
456 Several different clients should exist both to provide for a
457 variety of styles of human usage, and to support selection of
458 the most commonly used computer environments (e.g., UNIX, VMS,
459 MSDOS, OS2, MAC/OS).
460
461 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Pre-existing Services</span>
462
463 This section identifies other naming services that have been proposed
464 or implemented for naming people. Implementations of all of these
465 exist, although some are still only experimental.
466
467 Internet Domain Naming Service
468
469 The Internet Domain Name Service [6,1] is used today to name
470 host machines. It is implemented to address the query rates
471 and database sizes consistent with looking up hosts as part of
472 mail delivery. It provides a hierarchy with delegation of
473 authority within the hierarchy. Aliases are also available.
474 There is no access control, and the service is widely
475 distributed throughout the Internet. It supports management of
476 distribution, replication and caching. It is operational, and
477 provides a rich base of practical experience. It was
478 originally intended to be extensible to cover naming of people.
479 It runs on a variety of different operating systems and
480 utilizes the TCP/IP protocol suite.
481
482 The DECnet Network Architecture Naming Service (DNANS)
483
484 There is a rather well developed specification [5,3] for a
485 naming service that is part of the DECnet architecture, which
486 in turn arose from work at the DEC SRC in Palo Alto. This
487 architecture addresses some problems not yet covered by X.500,
488 such as access control, replication, and caching. It was
489 explicitly defined to have great scalability and management
490 features. It provides a global hierarchy of names, which are
491 mapped into properties. Therefore, operations of searching
492 based on properties or attributes may be expensive and
493 difficult. At present it is only implemented on VMS using the
494 DNA protocols, but will be moved to UNIX and TCP in the next
495 year.
496
497
498
499
500 <span class="grey">Sollins [Page 9]</span></pre>
501 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-10" ></span>
502 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
503
504
505 Clearinghouse
506
507 This service [7,2] is part of the Xerox network environment.
508 It operates today as a global service for Xerox. They have
509 considerable experience with its operation, including problems
510 of scale. Clearinghouse provides a three-level hierarchy of
511 names that are mapped to sets of properties. Loose consistency
512 is provided through slow propagation of updates. Both this
513 service and the DEC service mentioned above are to some extent
514 based on an earlier Xerox service called Grapevine.
515
516 Profile
517
518 A project at the University of Arizona run by Larry Peterson
519 [8] has produced a white pages name service called Profile. It
520 supports descriptive naming and sophisticated lookup tools.
521 Profile assumes the existence of some other service such as the
522 DNS to navigate among Profile servers. This navigation service
523 need not restrict the relationship among Profile servers to a
524 hierarchical organization; Profile supports a non-hierarchical
525 global structure. Names in Profile consist of sets of
526 attributes. Experimental implementations are in operation
527 today, and the largest site currently contains about 10,000
528 entries. The Profile code has been available for long enough
529 that it has become stable. The implementation is UNIX-based
530 only and uses TCP.
531
532 X.500
533
534 X.500 is the CCITT recommendation (also ISO/IEC/DIS 9594) [4]
535 for a directory service. Because it is a CCITT recommendation,
536 it evolves in four year study periods, one of which has
537 recently come to a close. Thus, X.500 has a stable definition
538 for the next four years.
539
540 In X.500, the set of all objects forms a single hierarchy, with
541 each object being named relative to its parent and a single
542 root as the topmost parent. An object consists of a set of
543 attributes. Searching can be done by use of a logical
544 combination of attribute values, known as a filter. A subset
545 of these attributes comprise an object's distinguished name
546 relative to its parent. The hierarchy as described in the
547 CCITT recommendation is geographic at its top level and
548 organizational within that. Alternatives can also be defined,
549 although they are not part of the CCITT or ISO documents. In
550 addition, there is no proposed mechanisms for distributing
551 information about other attribute types or object classes. As
552 with the other services, X.500 is a distributed service. It
553
554
555
556 <span class="grey">Sollins [Page 10]</span></pre>
557 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-11" ></span>
558 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
559
560
561 specifies cooperating servers or Directory Server Agents (DSAs)
562 under local control and management each of which knows about
563 one or more parts of the hierarchy. The clients are known as
564 Directory User Agents (DUAs). It is defined to run on top of
565 the OSI protocol stack. The demonstrations of X.500 in the
566 context of Internet run on top of the ISODE package, which
567 provides OSI transport on top of TCP.
568
569 X.500 is incomplete in that there are a number of identifiable
570 areas in which the standard says nothing, but that need to be
571 specified for a successful implementation. Some examples of
572 these are: access control (although authentication is
573 supported), replication, caching, the database itself (the
574 shape of the hierarchy), tools to limit the scope and cost of
575 searching, and database management tools.
576
577 There are currently a small number of implementations of X.500
578 in progress at such locations as University College London (the
579 Quipu project, on UNIX using ISODE), the University of British
580 Columbia (UNIX based using their own full OSI suite), MIT
581 (experimental, Symbolics Lisp Machine based, Lisp using TCP),
582 The Wollongong Group (offshoot of Quipu), The Retix
583 Corporation, NIST, and at least several underway in Italy and
584 Japan. There are probably others and a number of other
585 American corporations have discussed building their own. Each
586 of these must make its own decision in the areas in which X.500
587 is silent. Quipu is probably the most complete implementation
588 of X.500 to date. The pilot version has about 20 DUAs in seven
589 countries with an estimated 20,000 entries total.
590
591 <span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Proposed Approach</span>
592
593 The conclusion of this report is that some form of X.500 is the most
594 likely candidate. The reasons for this decision are that it has a
595 rich semantics and will become the international de facto standard.
596 There are, however, serious problems with its incompleteness and with
597 its strict hierarchy. Therefore, in order to explore these and
598 become convinced of its viability, the attendees at the meeting
599 agreed on field trials, as a first stage. Initially, this would
600 include experiments with at least one X.500 implementation (Quipu),
601 Profile to explore a non-hierarchical structure and richer
602 descriptive naming, and DNANS in order to explore some of the
603 incomplete aspects of X.500 for which DNANS has architected
604 solutions.
605
606 A three-stage plan, with all three stages beginning coincidentally
607 and as soon as possible, would provide such a service within the NRN.
608 The first stage should be complete in a year, the second in two, and
609
610
611
612 <span class="grey">Sollins [Page 11]</span></pre>
613 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-12" ></span>
614 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
615
616
617 the third in three. Stage 1 would be field trials of three
618 approaches to naming with an emphasis on distinguishing between the
619 specification and a particular implementation of X.500, as well.
620 Stage 2 would be a more complete implementation of a white pages
621 service base on the conclusions from Stage 1. Stage 3 would be
622 widespread deployment of the implementation developed in Stage 2.
623 The planning for Stage 3 is not outlined here in detail, because that
624 plan would be part of the proposed work to be done. If the field
625 trials were to lead to the conclusion that none of the services is
626 adequate, the plan for the remainder of the work would need to be
627 rescheduled.
628
629 If the Internet community is to adopt X.500 (or any other standard),
630 it is necessary to make a number of design and management decisions,
631 above and beyond the implementation decisions for the DSA. Since
632 there are a number of such decisions to be resolved, and some of
633 these are significant, the group recommended that this planning and
634 management function should be recognized as a distinct activity.
635
636 <span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. Stage 1: The Field Test</span>
637
638 It was agreed that field trials would be a valuable form in which to
639 explore the issues of building a white pages service for two reasons.
640 First, the software is still in early stages of development or
641 deployment. Some of it is production code, but still first release;
642 the rest is part of research projects. Second, it is important to
643 learn from experience with a limited and sympathetic community. The
644 suggested community was the computer science community, in
645 particular, computer science departments. That will not be the case
646 completely, since the computer science community in general does not
647 use DECnet. Therefore, for experiments with the DNANS, the NASA/DOE
648 community was recommended. They will be using DNANS in any case, as
649 they move to DECnet Phase V.
650
651 The twofold purpose of the field trials is to explore differing
652 directory service architectures and to refine the study of X.500
653 specifically, to distinguish architectural aspects of it from
654 features of a particular implementation of X.500. Initially, the
655 trials would include the Quipu implementation of X.500, Profile, and
656 the DNANS. A second implementation of X.500 should be identified and
657 included as soon as possible. Part of the emphasis of the field
658 trials would be on gathering and maintenance of naming information.
659 To ease this, a single common file format for storage of and access
660 to the naming information and use of a single set of data management
661 tools was recommended, although no particular set was identified.
662 The various directory services would need to be retrofitted to this
663 file format. Such consistency in file format would mean that the
664 services could all be co-resident, sharing files, thus permitting
665
666
667
668 <span class="grey">Sollins [Page 12]</span></pre>
669 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-13" ></span>
670 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
671
672
673 single locations to participate in several parts of the field trials.
674 This, in turn, would allow for direct comparisons.
675
676 There are a number of issues, which are not addressed in X.500, that
677 would need to be resolved for a large scale deployment such as a
678 white pages for the NRN. In particular, these are: clients of the
679 service; data collection and maintenance; distribution, replication
680 and caching of information; access control, accountability, and
681 information integrity; and support by non-OSI protocols. Each of the
682 name services included in the field trials would include decisions in
683 these areas, albeit different ones. The field trials will allow for
684 evaluation of these different mechanisms.
685
686 There are two other major issues that must also be addressed:
687 functionality and size. Functionality encompasses both the first
688 point of the nature of the interfaces to the service as well as the
689 structure of the namespace (e.g., hierarchy). A discussion of size
690 must include not only the number of entries handled by the service as
691 a whole, but how those entries are distributed and the query and
692 update patterns.
693
694 In general, all of these issues are tightly coupled, but are
695 separated here for the purposes of understanding the field trials and
696 its potential effectiveness. They would also be the issues that
697 would be the basis for the work done in Stage 2 of the project.
698
699 - Functionality:
700
701 X.500 and DNANS make strong statements about the organization
702 of the namespace. In both cases, it is a single, absolute
703 hierarchy with soft links or aliases and attribute-based naming
704 useful both in searches of subtrees of the hierarchy and for
705 storing information about the objects in the hierarchy. The
706 searches are based on logical combinations of attribute values.
707 Quipu implements the naming structure and search functionality
708 as specified in X.500. In contrast, Profile, provides a more
709 general facility that supports any form of relative names, not
710 just hierarchical, and a small programming language to express
711 the functions for searching. By including Profile in the field
712 trials, these more general facilities can be tested.
713
714 X.500 specifies that the service is separated into two parts
715 for implementation of the service, known as the Directory
716 Service Agent (DSA), and the client, known as the Directory
717 User Agent (DUA). DUAs can be implemented independently of the
718 implementation of the white pages service. Quipu, Profile, and
719 DNANS have taken different approaches to the presentation model
720 for DUAs, so the three implementations will allow for
721
722
723
724 <span class="grey">Sollins [Page 13]</span></pre>
725 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-14" ></span>
726 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
727
728
729 additional experience.
730
731 - Size:
732
733 As discussed earlier, a white pages service must be prepared to
734 handle a minimum of 10**7 entries, although they may be
735 distributed, and a query rate of hundreds per second. It must
736 also be prepared to handle much higher peak rates. If the
737 address lookup that is presently provided by the DNS is also
738 supported by the white pages service, the query rate will be
739 much higher. The designers of the field trials must determine
740 whether or not such usage will be part of the final service and
741 therefore must be examined in the field trials. If so, caching
742 may be part of the solution. In addition, the response time
743 for DUAs must be reasonable for a human sitting at a console.
744 Furthermore, modifications to the data should occur in
745 reasonably short periods of time, although this could be
746 measured in hours.
747
748 The field trials must allow for experimentation under such
749 stressful conditions. The environment for testing must have
750 both large and small nodes, as well as both heavy and light
751 load querying and situations in which reorganization can be
752 tested. Such reorganization may be a simple as moving one
753 piece of the hierarchy to another point and handling naming
754 conflicts in the new environment. X.500 does not address this
755 issue, but it will be needed by the NRN.
756
757 - Distribution, replication, and caching:
758
759 These are areas in which X.500 has very little to say, but a
760 great deal of work has been done in other distributed, network
761 naming services, in particular both the DNS and DNANS. There
762 seems to be general agreement that distribution of naming
763 services should be done on the basis of nodes in the naming
764 structure, which also provide the basis for administrative
765 partitioning. All the naming services described here support
766 distribution, partitioning of the information for placement on
767 cooperating servers. Neither X.500 (and therefore Quipu) nor
768 Profile is prepared to redistribute portions of the namespace,
769 for reallocation of administrative responsibilities or load
770 balancing, although this should be possible and DNANS is
771 prepared to do so. Replication is necessary for accessibility
772 in a large-scale or global namespace, although again X.500 does
773 not address this issue. Quipu has taken a stand on this, by
774 defining master and slave copies of the data; it is similar to,
775 but not the same as, the approach taken in the DNS. Caching is
776 barely touched on in X.500 and not at all in Profile, but our
777
778
779
780 <span class="grey">Sollins [Page 14]</span></pre>
781 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-15" ></span>
782 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
783
784
785 experience with the DNS indicates that caching is critical to
786 effective operation of a distributed name service. The DNANS
787 has an architected solution based on objects in the namespace
788 as the unit of distribution and replication. Again, the DNANS
789 solution should be explored in the field test environment.
790
791 - Access control, accountability, and integrity:
792
793 Access control and accountability require some degree of
794 authentication. X.500 supports authentication based on using
795 an RSA public key algorithm, but does not address issues of
796 universal registration, nor issues of access control or
797 accountability themselves. These are left as a local issue,
798 although depending on the design of the system, they may have
799 global implications. The problem of integrity of the
800 information in the name service is nowhere addressed. Profile
801 also does not address these issues, although it uses
802 authentication based on UNIX authentication, involving user ids
803 and passwords. DNANS takes a strong stand on access control,
804 architecting it in at the level of individual entries. Field
805 trials will force these issues out into the open.
806
807 - Structure of the naming tree:
808
809 In the deployment of the DNS, about one year was lost to
810 arguments about the actual structure of the naming hierarchy.
811 People form strong opinions about their name, and fight for or
812 against certain hierarchical structures. The same issue will
813 arise here, and advanced planning to deal with the problem is
814 required.
815
816 In this case, the problem is made harder by the fact that the
817 hierarchy will be global; X.500 is an international standard,
818 based on the assumption that there is only one example of the
819 tree, partitioned by country. Probably the American White
820 Pages Service, at least at its root, will be run by the NIST or
821 its contractor. We must deal with the problem that in the
822 short term, various implementations may not interwork, and we
823 must work with NIST to support the needed services.
824
825 Specific issues that come up related to the naming tree are:
826
827 * How is delegation of control of the tree managed?
828 For example, who decides what DSA holds what parts
829 of the tree?
830
831 * How is the creation of new parts of the tree
832 (e.g., an organizational entry) controlled?
833
834
835
836 <span class="grey">Sollins [Page 15]</span></pre>
837 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-16" ></span>
838 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
839
840
841 - Support for Tree Search:
842
843 Regardless of the defintion of the white pages service in the
844 NRN, it will need to interface to the X.500 world. The X.500
845 naming hierarchy can be expected to become very large, and
846 guidance is needed for users to help them navigate the tree.
847 Users need help to find their way to unknown parts of the
848 namespace. As in other naming services, a feature of X.500 is
849 that additional entries, aliases (similar to links in file
850 systems) can be installed to provide an easy path for a user in
851 one part of the tree to find other interesting parts of the
852 tree. By establishing a consistent policy for the use of alias
853 entries, learning how to navigate the tree can be made much
854 easier for a user. As part of setting up the tree, therefore,
855 these sorts of policies need to be defined.
856
857 - Definition of database structures:
858
859 There are a number of data structures that need to be defined
860 as part of setting up each of the services. These include, for
861 example, the types of information stored for the entry about a
862 person. This information must be stored in the servers, and
863 passed to the clients. These structures must thus be
864 specified. In other words, the schema defining attributes and
865 object classes must be specified for the NRN.
866
867 - Load balancing:
868
869 The dynamic performance of the Internet system must be
870 estimated, so that the servers can be sized properly.
871 Especially at the root of the tree, the query rate must be
872 estimated carefully. Caching will have a strong influence on
873 this. Therefore, traffic patterns are very dependent on the
874 details of implementation.
875
876 - Supporting multiple protocol suites:
877
878 At least three protocol suites are and will continue to be used
879 in the NRN environment. They are DECnet, TCP/IP, and the OSI
880 suite of protocols. Since the white pages service is at the
881 applications layer, it must run on top of at least these three
882 protocol suites. It is important to understand the
883 requirements of the white pages service for its transport
884 protocols.
885
886 By addressing these issues within the field trials, we will be
887 preparing for the further development of Stage 2. A result of Stage
888 1 will be a detailed specification of the white pages service,
889
890
891
892 <span class="grey">Sollins [Page 16]</span></pre>
893 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-17" ></span>
894 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
895
896
897 possibly an extension to or modification of X.500. This should
898 dovetail with the activities specifying the details required for
899 implementation (known as "profiling") by the NIST Workshop for
900 Implementors of OSI. In addition, in order to run the field trial,
901 the information capture problem must be addressed, providing the some
902 of the preliminary work of Stage 3.
903
904 <span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a>. Stage 2: Implementation</span>
905
906 If the evaluation of Stage 1 concludes that some form of X.500 is
907 acceptable, at least one of the two X.500 implementations included in
908 the field trials should provide the basis for a production quality
909 implementation of X.500 for general deployment. Further work will
910 likely be needed on the basis of the evaluations of the field trials.
911 A production version of an implementation requires both reliable
912 servers as well as a variety of clients to provide differing
913 interfaces on a mixture of hardware and operating systems.
914
915 In addition, especially because of the inclusion of Profile and
916 DNANS, a variety of different DUAs will be explored by definition.
917 Further investigation into the DUAs should begin in parallel with or
918 in conjunction with the field trials. There should be distinct DUAs
919 for both programs and humans. In addition, there probably should be
920 human-user DUAs geared both to the naive user with simple usage
921 patterns and the more sophisticated user who wants to perform complex
922 queries. It is also important to design DUAs that do not require a
923 great deal of computing power for the small machines still in use in
924 great quantity. Much of the user community may not be able to afford
925 expensive equipment upgrades.
926
927 Assuming that X.500 is deemed to be the specification of the service,
928 the field trials will address many issues not included in X.500 as of
929 1989. Since it is important for the NRN to support interconnectivity
930 beyond its own bounds, it behooves us to feed what has been learned
931 back into the standards activities. This was identified as a
932 separate activity because of the intellectual as well as time
933 commitment that must be made to do this effectively.
934
935 <span class="h3"><a class="selflink" id="section-4.3" href="#section-4.3">4.3</a>. Stage 3: Deployment</span>
936
937 A plan is required to develop the schedule of service introduction,
938 and to co-ordinate the deployment as it is undertaken. This includes
939 mediating service problems, a significant task in its own right.
940
941 The details of deployment were not discussed at the meeting, although
942 several of the seeds of deployment lie in Stages 1 and 2. The first
943 of these is the capture and management of information. The second is
944 DUA development. Both of these must be included Stage 1 in order to
945
946
947
948 <span class="grey">Sollins [Page 17]</span></pre>
949 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-18" ></span>
950 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
951
952
953 support a usable environment for the trials. In addition, the
954 information that will have been captured in Stage 1 could be printed
955 producing a hard copy of the white pages information. That could be
956 distributed to all scientists and engineers involved; such a project
957 would provide an early white pages service. During the initial
958 periods of both Stages 1 and 2, planning for deployment would also
959 have to proceed, in order to provide a smooth transition to this
960 third stage in the project.
961
962 <span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Conclusion</span>
963
964 The consensus of the meeting was that following a path that included
965 X.500 was both the correct direction and feasible, although X.500
966 needs further elaboration. There were several important items for
967 further study. The first is that there are many issues left
968 unresolved in X.500 that have been addressed in other naming
969 services, and the NRN should take advantage of the solutions in those
970 other services. The second is that there was some reservation about
971 certain features of X.500, especially in the area of the imposition
972 of a hierarchy for naming, and only limited flexibility in
973 descriptive naming. The participants believe that is important
974 understand whether X.500 provides enough mechanisms to work around
975 such problems by finding a higher common ground that includes the
976 best features of all the naming services included in the field
977 trials. The final issue with respect to X.500 was that there was
978 agreement that X.500 will be an accepted and utilized standard in at
979 least part of the networked community and therefore interfacing to it
980 will be necessary. Given that, and the other reasons for choosing
981 X.500, the consensus was that the plan described above would bring
982 the NRN and its community of users a useful and usable white pages
983 service.
984
985 References
986
987 1. Austein, R., "The Internet Domain Name System", Proceedings of
988 USA Decus, Massachusetts Institute Technology/LCS, 1987.
989
990 2. Demers, A., D. Greene, C. Hauser, W. Irish, J. Larson, S.
991 Shenker, H. Sturgis, D. Swinehart, and D. Terry, "Epidemic
992 algorithms for replicated database maintenance", Proceedings of
993 the 6th Symposium on Principles of Distributed Computing, ACM,
994 Vancouver, B.C., Canada, pp. 12-21, August 1987.
995
996 3. Digital Equipment Corporation, "DNA Naming Service Functional
997 Specification Version 1.0.1", Order number: EK-DNANS-FS-001,
998 Digital Equipment Corporation, 1988.
999
1000 4. International Organization for Standardization, "Information
1001
1002
1003
1004 <span class="grey">Sollins [Page 18]</span></pre>
1005 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-19" ></span>
1006 <span class="grey"><a href="./rfc1107">RFC 1107</a> A Plan for Internet Directory Services July 1989</span>
1007
1008
1009 Processing Systems - Open Systems Interconnection - The
1010 Directory", Draft Standard (In 8 parts), Also CCITT
1011 Recommendation X.500, 1988.
1012
1013 5. Lampson, B., "Desiging a Global Name Service," Proceedings of the
1014 5th Symposium on Principles of Distribute Computing, ACM,
1015 Calgary, Alberta, Canada, pp. 1-10, August 1986.
1016
1017 6. Mockapetris, P., "Domain Names - Concept and Facilities", <a href="./rfc1034">RFC</a>
1018 <a href="./rfc1034">1034</a>, USC/Information Sciences Institute, November 1987.
1019
1020 7. Oppen, D., and Y. Dalal, "The Clearinghouse: A Decentralized
1021 Agent for Locating Named Objects in a Distributed Environment",
1022 Tech. Rept. OPD-T8103, Xerox Corporation, Palo Alto, CA, October
1023 1981.
1024
1025 8. Peterson, L., "Profile: A System for Naming Internet Resources",
1026 Tech. Rept. 20, Department of Computer Science, University of
1027 Arizona, June 1987.
1028
1029 Author's Address
1030
1031 Karen R. Sollins
1032 Massachusetts Institute of Technology
1033 Laboratory for Computer Science
1034 545 Technology Square
1035 Cambridge, MA 02139-1986
1036
1037 Phone: (617) 253-6006
1038
1039 EMail: [email protected]
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060 Sollins [Page 19]
1061 </pre>