Debian Codegraph @

rfc1346.html

raw
1 <pre>Network Working Group P. Jones
2 Request for Comments: 1346 Joint Network Team, UK
3 June 1992
4
5
6 <span class="h1">Resource Allocation, Control, and Accounting</span>
7 <span class="h1">for the Use of Network Resources</span>
8
9 Status of this Memo
10
11 This memo provides information for the Internet community. It does
12 not specify an Internet standard. Distribution of this memo is
13 unlimited.
14
15 <span class="h2"><a class="selflink" id="section-0" href="#section-0">0</a>. MANAGEMENT SUMMARY</span>
16
17 This paper gives reasons for wanting better sharing mechanisms for
18 networks. It concludes that the challenge of sharing network
19 resources (and for example intercontinental link resources) between
20 groups of users is neither well understood, nor well catered for in
21 terms of tools for those responsible for managing the services. The
22 situation is compared with other fields, both inside and outside IT,
23 and examples are cited. Recommendations for further work are made.
24
25 The purpose of this RFC is to focus discussion on particular
26 challenges in large service networks in general, and the
27 International IP Internet in particular. No solution discussed in
28 this document is intended as a standard. Rather, it is hoped that a
29 general consensus will emerge as to the appropriate solutions,
30 leading eventually to the adoption of standards.
31
32 The structure of the paper is as follows:
33
34 1. Findings
35 2. Conclusions
36 3. Recommendations
37
38 <span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. FINDINGS</span>
39
40 Issues arising from contention in the use of networks are not
41 unusual. Once connectivity and reliability have been addressed to a
42 reasonable level, bandwidth becomes (or appears to become?) the main
43 issue. Usage appears to have a strong tendency to rise to fill the
44 resources available (fully in line with the principles of Parkinson's
45 Law). Line-speed upgrades have an effect, but with no guarantee of
46 permanently alleviating the problem. Line-speeds are increasing as
47 technology improves over time, but the variations on matters like
48 availability and funding are wide, and users remain avaricious.
49
50
51
52 <span class="grey">Jones [Page 1]</span></pre>
53 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span>
54 <span class="grey"><a href="./rfc1346">RFC 1346</a> Resource Allocation, Control, and Accounting June 1992</span>
55
56
57 Often the situation can appear worse than having to survive in a
58 jungle, in the sense that the strong (even if "good") seem to have
59 little advantage over the weak. It may seem that it is the
60 determined person rather than the important work that gets service.
61
62 Most people will have experienced poor service on an overloaded
63 network at some time. To help the end-users, it seems on the face of
64 it that one must help the IT Service Manager he relates to. Examples
65 relating to the relationship between the network manager and his
66 customers, IT Service Managers at institutions connecting to his
67 network, include the following:
68
69 (a) If the IT Service Manager finds his link to the Network Manager's
70 network overloaded, he may be offered a link upgrade, probably with a
71 cost estimate. He might prefer control mechanisms whereby he can say
72 that department X deserves more resources than department Y, or that
73 interactive terminal use takes preference over file transfers, or
74 that user U is more important than user V.
75
76 (b) Where an IT Service Manager is sharing a link, he will commonly
77 get more than his institution's share of the link, and often get very
78 good value-for-money compared to using a dedicated link, but he has
79 no guarantee that his end-users' usage won't get swamped by the use
80 of other (perhaps much larger) partners on the shared link. This
81 could be seen as wishing to have a guaranteed minimum share according
82 to some parameter(s).
83
84 (c) On a shared link as under (b), the Network Manager may wish to
85 ensure that usage of the link (which might be a high-performance
86 trunk line on a network or an international link for example) by any
87 one partner is "reasonable" in relation perhaps to his contribution
88 to the costs. In contrast to (b), the Network Manager is wishing to
89 impose a maximum value on some parameter(s). He may be happy if the
90 width of the IT Service Manager's access link is not greater than his
91 share of the shared link (assuming the measure agreed on is "width"),
92 but this will commonly not be the case. To be able to reach
93 agreement, the Network Manager and the IT Service Manager may need
94 options on the choice of parameters, and perhaps a choice on the
95 means of control, as well as being able to negotiate about values.
96
97 In circumstances where the Network Manager can exercise such controls
98 over his customers, the IT Service Managers may say with some feeling
99 and perhaps with justification, that if they are going to be
100 controlled can the Network Manager please provide tools whereby they
101 can arrange for the onward sharing of the resource they have, and
102 thence onwards down the hierarchy to the end-users.
103
104
105
106
107
108 <span class="grey">Jones [Page 2]</span></pre>
109 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span>
110 <span class="grey"><a href="./rfc1346">RFC 1346</a> Resource Allocation, Control, and Accounting June 1992</span>
111
112
113 (d) It may be Network Manager A has a link that Network Manager B
114 would like to use on occasion, perhaps as back-up on access to a
115 third network. Network Manager A might well wish to be
116 accommodating, perhaps as examples because of financial benefit or
117 perhaps because of the possibility of a reciprocal arrangement.
118 However, the fear of overload affecting normal use and the lack of
119 control over the usage militates against arrangements that the
120 parties could be quite keen to make.
121
122 Such challenges are very far from being unique to networking.
123 Government and both public and private organisations and companies
124 allocate budgets (and resources other than money), control and
125 account for usage, recognising the possibility of overdrawing and
126 borrowing. In times of shortage, food is rationed. I haven't
127 checked this out, but it would surprise me if Jerry Hall wasn't
128 guaranteed a ticket for any Rolling Stones concert, should she wish
129 to attend.
130
131 The charging factor influences use but does not control it (except
132 perhaps in unusual circumstances where say payment was expected in
133 advance and usage was cut off when the money ran out).
134
135 In the IT world, multi-user hosts have filestore control systems; one
136 that I use has an overdraft facility with no penalty for not having a
137 prior arrangement! There are also system designs and implementations
138 for sharing host processor time with more sophistication than just
139 counting seconds and chopping people off; this problem seems to me to
140 be reasonably well understood. (Library catalogue searches under
141 author "John Larmouth" should provide some references for those who
142 require convincing.) Some multi-user hosts have controls of sorts on
143 terminal connections. On the other hand, I am not aware of any
144 control system in operation that can guarantee multi-user host
145 response time even outside the network context among directly
146 connected terminals.
147
148 The various roles bring different interests to bear. A provider will
149 not necessarily see it in his interests to control usage, or (perhaps
150 even more likely) to provide customers with control tools, since the
151 lack of these may encourage - or even oblige - the customer to buy
152 more. Even if the IT Service Manager can deal with the issue of who
153 or what is important, and the issues of the relative importance of
154 allocating resources against requests, other issues like social
155 acceptability may arise to complicate his life. For example it may
156 be generally agreed (and perhaps the network manager instructed) that
157 "everyone" must be able to do a small amount of work at any time,
158 perhaps to do some housekeeping or seek information.
159
160
161
162
163
164 <span class="grey">Jones [Page 3]</span></pre>
165 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span>
166 <span class="grey"><a href="./rfc1346">RFC 1346</a> Resource Allocation, Control, and Accounting June 1992</span>
167
168
169 Time is an important factor. Network resources, like computer
170 processor time and unlike filestore, vanish if they are not used.
171 People will in general prefer resources during prime shift to those
172 in the middle of their night; however, in global terms the middle of
173 their night can be during prime shift somewhere along their path of
174 usage.
175
176 What's to do? Splitting lines with multiplexers is rather
177 inflexible, and may well militate against the benefits of resource-
178 sharing that give rise commonly to link-sharing arrangements. Some
179 technologies:
180
181 - have the ability to treat (or at least mark) traffic as of high
182 priority, for example where it gives emergency or status
183 information;
184
185 - (in the case of X.25(84), I understand from my JNT colleague Ian
186 Smith,) have throughput class (<a href="#section-6.13">section 6.13</a>) and transit delay
187 (<a href="#section-6.27">section 6.27</a>). (Ian tells me that it is in his view far from
188 clear how practical these facilities are);
189
190 - may be able to discriminate between traffic on grounds of
191 network source address;
192
193 - may be able to discriminate between traffic on grounds of
194 network destination address;
195
196 - may be able to discriminate between traffic on grounds of
197 application protocol, perhaps giving preference to interactive
198 terminal traffic, or making a choice between preference for
199 email and for file transfer traffic;
200
201 - may be able to discriminate between traffic on grounds of other
202 facets of network protocol or traffic.
203
204 In practice, one may well not have adequate tools in these or other
205 terms, and one may well have to ignore the challenges of resource
206 control, and either ignore the issue or refuse service.
207
208 <span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. CONCLUSIONS</span>
209
210 2.1 There seems to be a lack of tools to enable the controlling
211 and the sharing of networks and links. This is militating against
212 the cooperative sharing of resources, and restricting the ability
213 of organisations to do business with one another.
214
215 2.2 Further, the definition of what constitutes a share, or what
216 parameter of service one would try to measure and control (or what
217
218
219
220 <span class="grey">Jones [Page 4]</span></pre>
221 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span>
222 <span class="grey"><a href="./rfc1346">RFC 1346</a> Resource Allocation, Control, and Accounting June 1992</span>
223
224
225 the choices are if any), is not clear.
226
227 2.3 Following from that, it is then not clear whether what is
228 needed is new or enhanced protocols/services, new or enhanced
229 procurement specifications or profiles, or new or enhanced
230 networking products or tools.
231
232 2.4 Service providers (more likely the public carriers or but also
233 some Network Managers) may see it as against their interests to
234 provide controlling tools if they see them as tending to constrain
235 usage and hence reducing income. If so, they may not support, and
236 may even oppose, progress in the area. However, they might be
237 persuaded that the provision of such tools might give them
238 competitive edge over their rivals, and therefore to support
239 appropriate projects and developments.
240
241 <span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. RECOMMENDATIONS</span>
242
243 There seems scope for one or more studies to:
244
245 - restate and refine the definition of the problems;
246
247 - collect, catalogue and relate relevant experience in both the
248 networking and non-networking fields;
249
250 - make recommendations as to what areas (e.g., among those
251 suggested in 2.3 above) projects should be undertaken;
252
253 - outline possible projects, indicating the timescale on which
254 improved sharing of production network service resources is
255 likely to be achieved, and recommending an order of priority
256 among the suggested projects.
257
258 FOOTNOTES:
259
260 Gender issues - where appropriate, the male embraces the female and
261 vice versa.
262
263 Dramatis Personae:
264
265 Jerry Hall is a close associate of Mr. M. Jagger, formerly of the
266 London School of Economics in the University of London, and now
267 Chairman and Chief Executive of an internationally prominent and
268 successful commercial musical operation.
269
270 Others mentioned in this paper are assumed to prefer to remain
271 anonymous, although the standard is to give contact information
272 for the author (see Author's Address section).
273
274
275
276 <span class="grey">Jones [Page 5]</span></pre>
277 <hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span>
278 <span class="grey"><a href="./rfc1346">RFC 1346</a> Resource Allocation, Control, and Accounting June 1992</span>
279
280
281 Security Considerations
282
283 Security issues are not discussed in this memo.
284
285 Author's Address
286
287 Phil Jones
288 JNT
289 RAL, Chilton, Didcot, OXON OX11 0QX
290
291 Voice: +44-235-446618
292 Fax: +44-235-446251
293
294 Email: [email protected] or c=gb;a= ;p=uk.ac;o=jnt;i=p;s=jones;
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332 Jones [Page 6]
333 </pre>