idnits 2.17.00 (12 Aug 2021) /tmp/idnits47084/draft-nishida-sctp-failover-00.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 doesn't use any RFC 2119 keywords, yet seems to have RFC 2119 boilerplate text. -- The document date (December 13, 2009) is 4535 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) -- Possible downref: Non-RFC (?) normative reference: ref. 'CARO02' -- Possible downref: Non-RFC (?) normative reference: ref. 'CARO04' -- Possible downref: Non-RFC (?) normative reference: ref. 'CARO05' -- Possible downref: Non-RFC (?) normative reference: ref. 'FALLON08' -- Possible downref: Non-RFC (?) normative reference: ref. 'GRINNEMO04' -- Possible downref: Non-RFC (?) normative reference: ref. 'IYENGAR06' -- Possible downref: Non-RFC (?) normative reference: ref. 'JUNGMAIER02' -- Possible downref: Non-RFC (?) normative reference: ref. 'NATARAJAN08' ** Downref: Normative reference to an Informational RFC: RFC 4690 Summary: 2 errors (**), 0 flaws (~~), 2 warnings (==), 9 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group Y. Nishida 3 Internet-Draft WIDE Project 4 Intended status: Standards Track December 13, 2009 5 Expires: June 16, 2010 7 Quick Failover Algorithm in SCTP 8 draft-nishida-sctp-failover-00 10 Abstract 12 One of the major advantages in SCTP is supporting multi-homing 13 communication. If an multi-homed end-point has redundant network 14 connections, sctp sessions can have a good chance to survive from 15 network failures by migrating inactive network to active one. 16 However, if we follow the SCTP standard, there can be significant 17 delay for the network migration. During this migration period, SCTP 18 cannot transmit much data to the destination. This issue drastically 19 impairs the usability of SCTP in some situations. This memo 20 describes the issue of SCTP failover mechanism and discuss its 21 solutions which require minimal modification to the current standard. 23 Status of this Memo 25 This Internet-Draft is submitted to IETF in full conformance with the 26 provisions of BCP 78 and BCP 79. 28 Internet-Drafts are working documents of the Internet Engineering 29 Task Force (IETF), its areas, and its working groups. Note that 30 other groups may also distribute working documents as Internet- 31 Drafts. 33 Internet-Drafts are draft documents valid for a maximum of six months 34 and may be updated, replaced, or obsoleted by other documents at any 35 time. It is inappropriate to use Internet-Drafts as reference 36 material or to cite them other than as "work in progress." 38 The list of current Internet-Drafts can be accessed at 39 http://www.ietf.org/ietf/1id-abstracts.txt. 41 The list of Internet-Draft Shadow Directories can be accessed at 42 http://www.ietf.org/shadow.html. 44 This Internet-Draft will expire on June 16, 2010. 46 Copyright Notice 48 Copyright (c) 2009 IETF Trust and the persons identified as the 49 document authors. All rights reserved. 51 This document is subject to BCP 78 and the IETF Trust's Legal 52 Provisions Relating to IETF Documents 53 (http://trustee.ietf.org/license-info) in effect on the date of 54 publication of this document. Please review these documents 55 carefully, as they describe your rights and restrictions with respect 56 to this document. Code Components extracted from this document must 57 include Simplified BSD License text as described in Section 4.e of 58 the Trust Legal Provisions and are provided without warranty as 59 described in the BSD License. 61 Table of Contents 63 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 64 2. Conventions and Terminology . . . . . . . . . . . . . . . . . 4 65 3. Issue in SCTP Path Management Process . . . . . . . . . . . . 5 66 4. Solutions for Smooth Failover . . . . . . . . . . . . . . . . 6 67 4.1. Reduce Path.Max.Retrans . . . . . . . . . . . . . . . . . 6 68 4.2. Introduce Potential Failure Status in Failure 69 Detection Algorithm . . . . . . . . . . . . . . . . . . . 7 70 5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 9 71 5.1. Effect of Path Bouncing . . . . . . . . . . . . . . . . . 9 72 5.2. Permanent Failover . . . . . . . . . . . . . . . . . . . . 9 73 6. Security Considerations . . . . . . . . . . . . . . . . . . . 10 74 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 11 75 8. Normative References . . . . . . . . . . . . . . . . . . . . . 12 76 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 13 78 1. Introduction 80 Multihoming support is one of the major advantage of SCTP which is 81 not supported in other transport protocols such as TCP or UDP. If an 82 multi-homed end-point has redundant network connections, SCTP 83 sessions can survive from the network failures by migrating inactive 84 path to active one. This feature can be expected to be a driving 85 force for deploying SCTP, however, because of minor issues in the 86 SCTP specification, most of SCTP sessions will have significant delay 87 to failover and will cause significant performance degradation during 88 the failover process. We believe this issue is impairing the 89 usability of SCTP and it is important to address it to make SCTP more 90 efficient and attractive. 92 In this memo, we describes the issue of SCTP failover process and 93 discuss the solutions. Our main focus is to propose a solution that 94 does not require major modification to the current standard. Using 95 Concurrent Multipath Transfer (CMT) [IYENGAR06] which allows SCTP 96 utilize multiple path simultaneously for data transmission can be an 97 another approach to solve this issue. CMT with sophisticated multi- 98 homing communication control may bring the ideal solution, however, 99 it might require to add a lot of additional functions to the current 100 standard. In addition, some may not want concurrent data transfer 101 feature, but want to use smooth failover feature in SCTP. From this 102 reason, we believe the proposals in this document can be useful and 103 meaningful. 105 2. Conventions and Terminology 107 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 108 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 109 document are to be interpreted as described in [RFC2119]. 111 Since this document describes a potential risk in NewReno, it uses 112 the same terminology and definitions in RFC4690. [RFC4690]. 114 3. Issue in SCTP Path Management Process 116 SCTP can utilize multiple IP addresses for single SCTP association. 117 Each SCTP endpoint exchanges the list of available addresses on the 118 node during initial negotiation. After this, endpoints select one 119 address from the list and define this as the destination of the 120 primary path. Basically, SCTP sends all data through this primary 121 path for normal data transmissions. Also, it sends heartbeat packets 122 to other (non-primary) destinations at a certain interval to check 123 the reachability of the path. 125 If sender has multiple active destination addresses, it can 126 retransmit data to non-primary destination address when the 127 transmission to the primary times out. 129 When sender receives the acknowledgment for data or heartbeat packets 130 from one of the destination addresses, it considers the destination 131 is active. If it fails to receive acknowledgments, the error count 132 for the address is increased. If the error counter exceeds the 133 protocol parameter 'Path.Max.Retrans', SCTP endpoint considers the 134 address is inactive. 136 The failover process of SCTP is initiated when the primary path 137 becomes inactive (error counter for the primacy path exceeds 138 Path.Max.Retrans). If the primary path is marked inactive, SCTP 139 chooses new destination address from one of the active destinations 140 and start using this address to send data. If the primary path 141 becomes active again, SCTP uses the primary destination for 142 subsequent data transmissions and stop using non-primary one. 144 An issue in this failover process is that it usually takes 145 significant amount of time before SCTP switches to the new 146 destination. Let's say the primary path on a multi-homed host 147 becomes unavailable and the RTO value for the primary path at that 148 time is around 1 second, it usually takes over 30 seconds before SCTP 149 starts to use the secondary path. This is because the recommended 150 value for Path.Max.Retrans in the standard is 5, which requires 5 151 consecutive timeouts before failover takes place Before SCTP switches 152 to the secondary address, SCTP keeps trying to send packets to the 153 primary and only retransmitted packets sent to the secondary can be 154 reached at the receiver. This slow failover process can cause 155 significant performance degradation and will not be acceptable in 156 some situations. 158 4. Solutions for Smooth Failover 160 The following approach are conceivable for the solutions of this 161 issue. 163 4.1. Reduce Path.Max.Retrans 165 If we choose smaller value for Path.Max.Retrans, we can shorten the 166 duration of failover process. In fact, this is recommended in some 167 research results [JUNGMAIER02] [GRINNEMO04] [FALLON08]. For example, 168 if we set Path.Max.Retrans to 0, SCTP switches to another destination 169 on a single timeout. However, smaller value for Path.Max.Retrans 170 might cause spurious failover which might cause bouncing paths. In 171 addition, if we use smaller value for Path.Max.Retrans, we may also 172 need to choose smaller value for 'Association.Max.Retrans'. The 173 Association.Max.Retrans indicates the threshold for the total number 174 of consecutive error count for the entire SCTP association. If the 175 total of the error count for all paths exceeds this value, the 176 endpoint considers the peer endpoint unreachable and terminates the 177 association. According to the Section 8.2 in RFC4960, we should 178 avoid having the value of Association.Max.Retrans larger than the 179 summation of the Path.Max.Retrans of all the destination addresses. 180 Otherwise, even if all the destination addresses become inactive, the 181 endpoint still considers the peer endpoint reachable. The behavior 182 in this situation is not defined in the RFC and depends on each 183 implementation. In order to avoid inconsistent behavior between 184 implementations, we had better use smaller value for 185 Association.Max.Retrans. However, if we choose smaller value for 186 Association.Max.Retrans, associations will prone to be terminated 187 with minor congestion. 189 Another issue is that the interval of heartbeat packet: 'HB.interval' 190 may not be small. (recommended value is 30 seconds) This means once 191 failover takes place, an endpoint might need a certain amount of time 192 to use the primary path again. This can cause undesirable effects in 193 case of spurious failover. If we choose smaller value for 194 HB.interval, the traffic used for path probing in a session will be 195 increased. 197 The advantage of turning Path.Max.Retrans is that it requires no 198 modification to the current standard, although it needs to ignore 199 several recommendations. In addition, some research results indicate 200 path bouncing caused by spurious failover does not cause serious 201 problems. We discuss the effect of path bouncing in the section 5. 203 4.2. Introduce Potential Failure Status in Failure Detection Algorithm 205 As seen above, one difficulty of tuning Path.Max.Retrans is that it 206 is required to meet the following two inconsistent requirements. 208 o In order to respond network failure quickly, we need to mark a 209 path as inactive as soon as we detect failure. 211 o In order to make an association persistent and robust against 212 network failure, we need to be conservative to mark a path as 213 inactive. 215 To satisfy these requirements, we propose to introduce "Potential 216 Failure" state in failure detection algorithm in SCTP. Potential 217 Failure state is the intermediate state between Active and Inactive. 218 It indicates that the path is possibly inactive, but not confirmed 219 yet. By using this state, SCTP can respond network failure quickly, 220 while it can preserve a conservative policy of marking path as 221 inactive. The idea of using Potential Failure state is originally 222 proposed in [NATARAJAN08] for CMT. 224 In this algorithm, when sender receives the acknowledgment for data 225 or heartbeat packets from one of the destination addresses, it 226 considers the destination is Active. If it fails to receive 227 acknowledgments, SCTP endpoint increment the error count for the path 228 and marks the path as Potential Failure. (we might need to have new 229 threshold value for error counter to be conservative to migrate from 230 Active to Potential Failure. But, we choose this way for now) 232 If the primary path is marked Potential Failure, SCTP chooses new 233 destination address from one of the active destinations and start 234 using this address to send data. SCTP endpoints should not send any 235 data packet to paths in Potential Failure state, however, it can send 236 heartbeat packets at a certain interval. To allow quick recover from 237 Potential Failure state, we also propose to introduce a new protocol 238 parameter 'PFHB.Interval'. PFHB.interval is used to determine the 239 interval of heartbeat packets. It is recommended that a heartbeat 240 packet is sent once per RTO of each destination address plus 241 PFHB.interval with jittering of +/- 50% of the RTO value. It is also 242 recommended to use relatively smaller value than HB.interval for 243 PFHB.interval. 245 If the heartbeat is answered, SCTP marks the path Active again. If 246 unanswered, SCTP increments the error count and use an exponential 247 backoff algorithm to increase the RTO. If the error count exceeds 248 Path.Max.Retrans, the path is marked as Inactive. If all paths 249 become Potential Failure state, SCTP endpoint should not send any 250 data to its peer, while it can send heartbeat packets. Except the 251 use of PFHB.interval, other rules of sending heartbeats are 252 completely the same as those of the standard. 254 The advantage of this approach is that we can keep the same values 255 for Path.Max.Retrans, Association.Max.Retrans and HB.interval used in 256 the current implementations, while it can respond network failure 257 quickly. In addition, new transmission algorithm becomes effective 258 only when the path is in Potential Failure state. When the primary 259 path is in Active or Inactive, the behavior is completely the same as 260 that of the current standard. Hence, the influences of the algorithm 261 will be limited. 263 5. Discussion 265 5.1. Effect of Path Bouncing 267 The methods described above can accelerate failover process. Hence, 268 it might introduce path bouncing effect which keeps changing the data 269 transmission path frequently. This sounds harmful for data transfer, 270 however several research results indicate that there is no serious 271 problem with SCTP in terms of path bouncing effect [CARO04] [CARO05]. 273 There are two main reasons for this. First, SCTP is basically 274 designed for multipath communication, which means SCTP maintains all 275 path related parameters (cwnd, ssthresh, RTT, error count, etc) per 276 each destination address. These parameters cannot be affected by 277 path bouncing. In addition, when SCTP migrates to another path, it 278 starts with minimal cwnd because of slow-start. Hence, there is 279 little chance for packet reordering or duplicating. 281 Second, even if all communication paths between end-nodes share the 282 same bottleneck, the proposed method does not make situations worse. 283 In case of congestion, the current standard tries to transmit data 284 packets to the primary during failover, while the proposed method 285 tries to explore other destinations. In any case, the same amount of 286 data packets sent to the same bottleneck. 288 5.2. Permanent Failover 290 When primary path becomes active again after failover, SCTP migrates 291 back to the primary path. After this, SCTP starts data transfer with 292 minimal cwnd. This is because SCTP must perform slow-start when it 293 migrates to new path. However, this might degrade the communication 294 performance in case that the performance of the alternative path is 295 relatively good. In order to mitigate this effect of slow-start, 296 permanent failover was proposed in [CARO02]. Permanent failover 297 allows SCTP to remain the alternative path even if the primacy path 298 becomes active again. This approach can improve performance in some 299 cases, however, it will require more detail analysis since it might 300 impact on SCTP failover algorithm. Since we prefer to keep the 301 current behavior of the standard as possible, we recommend not to 302 take this approach for now. 304 6. Security Considerations 306 There are no new security considerations introduced in this document. 308 7. IANA Considerations 310 This document does not create any new registries or modify the rules 311 for any existing registries managed by IANA. 313 8. Normative References 315 [CARO02] Caro Jr., A., Iyengar, J., Amer, P., Heinz, G., and R. 316 Stewart, "A Two-level Threshold Recovery Mechanism for 317 SCTP", Tech report, CIS Dept, University of Delaware , 318 7 2002. 320 [CARO04] Caro Jr., A., Amer, P., and R. Stewart, "End-to-End 321 Failover Thresholds for Transport Layer Multihoming", 322 MILCOM 2004 , 11 2004. 324 [CARO05] Caro Jr., A., "End-to-End Fault Tolerance using Transport 325 Layer Multihoming", Ph.D Thesis, University of Delaware , 326 1 2005. 328 [FALLON08] 329 Fallon, S., Jacob, P., Qiao, Y., Murphy, L., Fallon, E., 330 and A. Hanley, "SCTP Switchover Performance Issues in WLAN 331 Environments", IEEE CCNC 2008, 1 2008. 333 [GRINNEMO04] 334 Grinnemo, K-J. and A. Brunstrom, "Peformance of SCTP- 335 controlled failovers in M3UA-based SIGTRAN networks", 336 Advanced Simulation Technologies Conference , 4 2004. 338 [IYENGAR06] 339 Iyengar, J., Amer, P., and R. Stewart, "Concurrent 340 Multipath Transfer using SCTP Multihoming over Independent 341 End-to-end Paths.", IEEE/ACM Trans on Networking 14(5), 342 10 2006. 344 [JUNGMAIER02] 345 Jungmaier, A., Rathgeb, E., and M. Tuexen, "On the use of 346 SCTP in failover scenrarios", World Multiconference on 347 Systemics, Cybernetics and Informatics , 7 2002. 349 [NATARAJAN08] 350 Natarajan, P., Ekiz, N., Iyengar, J., Amer, P., and R. 351 Stewart, "Concurrent Multipath Transfer using SCTP 352 Multihoming: Introducing Potentially-failed Destination 353 State", IFIP Networking , 5 2008. 355 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 356 Requirement Levels", BCP 14, RFC 2119, March 1997. 358 [RFC4690] Klensin, J., Faltstrom, P., Karp, C., and IAB, "Review and 359 Recommendations for Internationalized Domain Names 360 (IDNs)", RFC 4690, September 2006. 362 Author's Address 364 Yoshifumi Nishida 365 WIDE Project 366 Endo 5322 367 Fujisawa, Kanagawa 252-8520 368 Japan 370 Email: nishida@wide.ad.jp