idnits 2.17.1 draft-dreibholz-rserpool-delay-05.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Sep 2009 rather than the newer Notice from 28 Dec 2009. (See https://trustee.ietf.org/license-info/) Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document 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 (January 5, 2010) is 5226 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 4960 (Obsoleted by RFC 9260) == Outdated reference: A later version (-32) exists of draft-ietf-tsvwg-sctpsocket-19 == Outdated reference: A later version (-34) exists of draft-dreibholz-rserpool-asap-hropt-04 == Outdated reference: A later version (-31) exists of draft-dreibholz-rserpool-enrp-takeover-01 Summary: 2 errors (**), 0 flaws (~~), 4 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group T. Dreibholz 3 Internet-Draft University of Duisburg-Essen 4 Intended status: Experimental X. Zhou 5 Expires: July 9, 2010 Hainan University 6 January 5, 2010 8 Definition of a Delay Measurement Infrastructure and Delay-Sensitive 9 Least-Used Policy for Reliable Server Pooling 10 draft-dreibholz-rserpool-delay-05.txt 12 Abstract 14 This document contains the definition of a delay measurement 15 infrastructure and a delay-sensitive Least-Used policy for Reliable 16 Server Pooling. 18 Status of this Memo 20 This Internet-Draft is submitted to IETF in full conformance with the 21 provisions of BCP 78 and BCP 79. 23 Internet-Drafts are working documents of the Internet Engineering 24 Task Force (IETF), its areas, and its working groups. Note that 25 other groups may also distribute working documents as Internet- 26 Drafts. 28 Internet-Drafts are draft documents valid for a maximum of six months 29 and may be updated, replaced, or obsoleted by other documents at any 30 time. It is inappropriate to use Internet-Drafts as reference 31 material or to cite them other than as "work in progress." 33 The list of current Internet-Drafts can be accessed at 34 http://www.ietf.org/ietf/1id-abstracts.txt. 36 The list of Internet-Draft Shadow Directories can be accessed at 37 http://www.ietf.org/shadow.html. 39 This Internet-Draft will expire on July 9, 2010. 41 Copyright Notice 43 Copyright (c) 2010 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 59 1.1. Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 60 1.2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . 3 61 1.3. Conventions . . . . . . . . . . . . . . . . . . . . . . . . 3 62 2. Delay-Measurement Infrastructure . . . . . . . . . . . . . . . 3 63 2.1. Quantification of Distance . . . . . . . . . . . . . . . . 3 64 2.2. Distance Measurement Environment . . . . . . . . . . . . . 4 65 3. Distance-Sensitive Least-Used Policy . . . . . . . . . . . . . 4 66 3.1. Description . . . . . . . . . . . . . . . . . . . . . . . . 5 67 3.2. ENRP Server Considerations . . . . . . . . . . . . . . . . 5 68 3.3. Pool User Considerations . . . . . . . . . . . . . . . . . 5 69 3.4. Pool Member Selection Policy Parameter . . . . . . . . . . 5 70 4. Reference Implementation . . . . . . . . . . . . . . . . . . . 6 71 5. Security Considerations . . . . . . . . . . . . . . . . . . . . 6 72 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 6 73 7. References . . . . . . . . . . . . . . . . . . . . . . . . . . 6 74 7.1. Normative References . . . . . . . . . . . . . . . . . . . 6 75 7.2. Informative References . . . . . . . . . . . . . . . . . . 7 76 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 8 78 1. Introduction 80 Reliable Server Pooling defines protocols for providing highly 81 available services. PEs of a pool may be distributed over a large 82 geographical area, in order to provide redundancy in case of 83 localized disasters. But the current pool policies defined in 84 [RFC5356] do not incorporate the fact of distances (i.e. delay) 85 between PU and PE. This leads to a low performance for delay- 86 sensitive applications. 88 1.1. Scope 90 This draft defines a delay measurement infrastructure for ENRP 91 servers to add delay information into the handlespace. Furthermore, 92 a delay-sensitive Least-Used policy is defined. Performance 93 evaluations can be found in [KiVS2007]. 95 1.2. Terminology 97 The terms are commonly identified in related work and can be found in 98 the Aggregate Server Access Protocol and Endpoint Handlespace 99 Redundancy Protocol Common Parameters document [RFC5354]. 101 1.3. Conventions 103 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 104 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 105 document are to be interpreted as described in [RFC2119]. 107 2. Delay-Measurement Infrastructure 109 This section describes the necessary delay measurement infrastructure 110 for the policy later defined in Section 3. It has to be provided as 111 part of the ENRP servers. 113 2.1. Quantification of Distance 115 Measuring delay for SCTP associations is easy: the SCTP protocol 116 [RFC4960] already calculates a smoothed round-trip time (RTT) for the 117 primary path. This RTT only has to be queried via the standard SCTP 118 API as defined in [I-D.ietf-tsvwg-sctpsocket]. By default, the 119 calculated RTT has a small restriction: a SCTP endpoint waits up to 120 200ms before acknowledging a packet, in order to piggyback the 121 acknowledgement chunk with payload data. In this case, the RTT would 122 include this latency. Using the option SCTP_DELAYED_ACK_TIME (see 123 [I-D.ietf-tsvwg-sctpsocket]), the maximum delay before acknowledging 124 a packet can be set to 0ms (i.e. "acknowledge as soon as possible"). 126 After that, the RTT approximately consists of the network latency 127 only. Then, using the RTT, the end-to-end delay between two 128 associated components is approximately 0.5*RTT. 130 In real networks, there may be negligible delay differences: for 131 example, the delay between a PU and PE #1 is 5ms and the latency 132 between the PU and PE #2 is 6ms. From the service user's 133 perspective, such minor delay differences may be ignored and are 134 furthermore unavoidable in Internet scenarios. Therefore, the 135 distance parameter between two components A and B is defined as 136 follows: 138 Distance = DistanceStep * round( (0.5*RTT) / DistanceStep ) 140 That is, the distance parameter is defined as the nearest integer 141 multiple of the constant DistanceStep for the measured delay (i.e. 142 0.5*RTT). 144 2.2. Distance Measurement Environment 146 In order to define a distance-aware policy, it is first necessary to 147 define a basic rule: PEs and PUs choose "nearby" ENRP servers. Since 148 the operation scope of RSerPool is restricted to a single 149 organization, this condition can be met easily by appropriately 150 locating ENRP servers. 152 o A Home ENRP server can measure the delay of the ASAP associations 153 to its PE. As part of its ENRP updates to other ENRP servers, it 154 can report this measured delay together with the PE information. 156 o A non-Home-ENRP server receiving such an update simply adds the 157 delay of the ENRP association with the Home ENRP server to the 158 PE's reported delay. 160 Now, each ENRP server can approximate the distance to every PE in the 161 operation scope using the equation in Section 2.1. 163 Note, that delay changes are propagated to all ENRP servers upon PE 164 re-registrations, i.e. the delay information (and the approximated 165 distance) dynamically adapts to the state of the network. 167 3. Distance-Sensitive Least-Used Policy 169 In this section, a distance-sensitive Least Used policy is defined, 170 based on the delay-measurement infrastructure introduced in 171 Section 2. 173 3.1. Description 175 The Least Used with Distance Penalty Factor (LU-DPF) policy uses load 176 information provided by the pool elements to select the lowest-loaded 177 pool elements within the pool. If there are multiple elements having 178 lowest load, the nearest PE should be chosen. 180 3.2. ENRP Server Considerations 182 The ENRP server SHOULD select at most the requested number of pool 183 elements. Their load values SHOULD be the lowest possible ones 184 within the pool and their distances also SHOULD be lowest. Each 185 element MUST NOT be reported more than once to the pool user. If 186 there is a choice of equal-loaded and equal-distanced pool elements, 187 round robin selection SHOULD be made among these elements. The 188 returned list of pool elements MUST be sorted by load value in 189 ascending order (1st key) and distance in ascending order (2nd key). 191 3.3. Pool User Considerations 193 The pool user should try to use the pool elements returned from the 194 list in the order returned by the ENRP server. A subsequent call for 195 handle resolution may result in the same list. Therefore, it is 196 RECOMMENDED for a pool user to request multiple entries in order to 197 have a sufficient amount of feasible backup entries available. 199 3.4. Pool Member Selection Policy Parameter 201 0 1 2 3 202 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 203 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 204 | Parameter Type = 0x6 | Length = 0x14 | 205 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 206 | Policy Type = 0x40000010 | 207 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 208 | Load | 209 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 210 | Load DPF | 211 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 212 | Distance | 213 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 215 o Load: Current load of the pool element. 217 o Load DPF: The LoadDPF setting of the PE. 219 o Distance: The approximated distance in milliseconds. 221 * Between PE and Home ENRP server: The distance SHOULD be set to 222 0. 224 * Between Non-Home ENRP server and Home ENRP server: The delay 225 measured on the ASAP association between Home ENRP server and 226 PE. 228 * Between ENRP server and PU: The sums of the measured delays on 229 the ASAP association and the ENRP association to the Home ENRP 230 server. 232 4. Reference Implementation 234 The RSerPool reference implementation RSPLIB can be found at 235 [RSerPoolPage]. It supports the functionalities defined by 236 [RFC5351], [RFC5352], [RFC5353], [RFC5354] and [RFC5356] as well as 237 the options [I-D.dreibholz-rserpool-asap-hropt], 238 [I-D.dreibholz-rserpool-enrp-takeover] and of course the option 239 defined by this document. An introduction to this implementation is 240 provided in [Dre2006]. 242 5. Security Considerations 244 Security considerations for RSerPool systems are described by 245 [RFC5355]. 247 6. IANA Considerations 249 This document does not require additional IANA actions beyond those 250 already identified in the ENRP and ASAP protocol specifications. 252 7. References 254 7.1. Normative References 256 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 257 Requirement Levels", BCP 14, RFC 2119, March 1997. 259 [RFC5351] Lei, P., Ong, L., Tuexen, M., and T. Dreibholz, "An 260 Overview of Reliable Server Pooling Protocols", RFC 5351, 261 September 2008. 263 [RFC5352] Stewart, R., Xie, Q., Stillman, M., and M. Tuexen, 264 "Aggregate Server Access Protocol (ASAP)", RFC 5352, 265 September 2008. 267 [RFC5353] Xie, Q., Stewart, R., Stillman, M., Tuexen, M., and A. 268 Silverton, "Endpoint Handlespace Redundancy Protocol 269 (ENRP)", RFC 5353, September 2008. 271 [RFC5354] Stewart, R., Xie, Q., Stillman, M., and M. Tuexen, 272 "Aggregate Server Access Protocol (ASAP) and Endpoint 273 Handlespace Redundancy Protocol (ENRP) Parameters", 274 RFC 5354, September 2008. 276 [RFC5355] Stillman, M., Gopal, R., Guttman, E., Sengodan, S., and M. 277 Holdrege, "Threats Introduced by Reliable Server Pooling 278 (RSerPool) and Requirements for Security in Response to 279 Threats", RFC 5355, September 2008. 281 [RFC5356] Dreibholz, T. and M. Tuexen, "Reliable Server Pooling 282 Policies", RFC 5356, September 2008. 284 [RFC4960] Stewart, R., "Stream Control Transmission Protocol", 285 RFC 4960, September 2007. 287 7.2. Informative References 289 [I-D.ietf-tsvwg-sctpsocket] 290 Stewart, R., Poon, K., Tuexen, M., Yasevich, V., and P. 291 Lei, "Sockets API Extensions for Stream Control 292 Transmission Protocol (SCTP)", 293 draft-ietf-tsvwg-sctpsocket-19 (work in progress), 294 February 2009. 296 [KiVS2007] 297 Dreibholz, T. and E. Rathgeb, "On Improving the 298 Performance of Reliable Server Pooling Systems for 299 Distance-Sensitive Distributed Applications", 300 Proceedings of the 15. ITG/GI Fachtagung Kommunikation in 301 Verteilten Systemen, February 2007. 303 [Dre2006] Dreibholz, T., "Reliable Server Pooling -- Evaluation, 304 Optimization and Extension of a Novel IETF Architecture", 305 Ph.D. Thesis University of Duisburg-Essen, Faculty of 306 Economics, Institute for Computer Science and Business 307 Information Systems, URL: http:// 308 duepublico.uni-duisburg-essen.de/servlets/DerivateServlet/ 309 Derivate-16326/Dre2006-final.pdf, March 2007. 311 [RSerPoolPage] 312 Dreibholz, T., "Thomas Dreibholz's RSerPool Page", 313 URL: http://tdrwww.iem.uni-due.de.de/dreibholz/rserpool/. 315 [I-D.dreibholz-rserpool-asap-hropt] 316 Dreibholz, T., "Handle Resolution Option for ASAP", 317 draft-dreibholz-rserpool-asap-hropt-04 (work in progress), 318 January 2009. 320 [I-D.dreibholz-rserpool-enrp-takeover] 321 Dreibholz, T. and X. Zhou, "Takeover Suggestion Flag for 322 the ENRP Handle Update Message", 323 draft-dreibholz-rserpool-enrp-takeover-01 (work in 324 progress), January 2009. 326 Authors' Addresses 328 Thomas Dreibholz 329 University of Duisburg-Essen, Institute for Experimental Mathematics 330 Ellernstrasse 29 331 45326 Essen, Nordrhein-Westfalen 332 Germany 334 Phone: +49-201-1837637 335 Fax: +49-201-1837673 336 Email: dreibh@iem.uni-due.de 337 URI: http://www.iem.uni-due.de/~dreibh/ 339 Xing Zhou 340 Hainan University, College of Information Science and Technology 341 Renmin Avenue 58 342 570228 Haikou, Hainan 343 China 345 Phone: +86-898-66279141 346 Email: zhouxing@hainu.edu.cn