idnits 2.17.1 draft-ietf-v6ops-pmtud-ecmp-problem-02.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. 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 date (June 17, 2015) is 3235 days in the past. Is this intentional? -- Found something which looks like a code comment -- if you have code sections in the document, please surround them with '' and '' lines. Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'Ether' is mentioned on line 234, but not defined Summary: 0 errors (**), 0 flaws (~~), 2 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 v6ops M. Byerly 3 Internet-Draft Fastly 4 Intended status: Informational M. Hite 5 Expires: December 19, 2015 Evernote 6 J. Jaeggli 7 Fastly 8 June 17, 2015 10 Close encounters of the ICMP type 2 kind (near misses with ICMPv6 PTB) 11 draft-ietf-v6ops-pmtud-ecmp-problem-02 13 Abstract 15 This document calls attention to the problem of delivering ICMPv6 16 type 2 "Packet Too Big" (PTB) messages to the intended destination in 17 ECMP load balanced or anycast network architectures. It discusses 18 operational mitigations that can be employed to address this class of 19 failure. 21 Status of This Memo 23 This Internet-Draft is submitted in full conformance with the 24 provisions of BCP 78 and BCP 79. 26 Internet-Drafts are working documents of the Internet Engineering 27 Task Force (IETF). Note that other groups may also distribute 28 working documents as Internet-Drafts. The list of current Internet- 29 Drafts is at http://datatracker.ietf.org/drafts/current/. 31 Internet-Drafts are draft documents valid for a maximum of six months 32 and may be updated, replaced, or obsoleted by other documents at any 33 time. It is inappropriate to use Internet-Drafts as reference 34 material or to cite them other than as "work in progress." 36 This Internet-Draft will expire on December 19, 2015. 38 Copyright Notice 40 Copyright (c) 2015 IETF Trust and the persons identified as the 41 document authors. All rights reserved. 43 This document is subject to BCP 78 and the IETF Trust's Legal 44 Provisions Relating to IETF Documents 45 (http://trustee.ietf.org/license-info) in effect on the date of 46 publication of this document. Please review these documents 47 carefully, as they describe your rights and restrictions with respect 48 to this document. Code Components extracted from this document must 49 include Simplified BSD License text as described in Section 4.e of 50 the Trust Legal Provisions and are provided without warranty as 51 described in the Simplified BSD License. 53 Table of Contents 55 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 56 2. Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 57 3. Mitigation . . . . . . . . . . . . . . . . . . . . . . . . . 4 58 3.1. Alternatives . . . . . . . . . . . . . . . . . . . . . . 5 59 3.2. Implementation . . . . . . . . . . . . . . . . . . . . . 5 60 3.2.1. Alternatives . . . . . . . . . . . . . . . . . . . . 6 61 4. Improvements . . . . . . . . . . . . . . . . . . . . . . . . 7 62 5. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 7 63 6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 7 64 7. Security Considerations . . . . . . . . . . . . . . . . . . . 7 65 8. Informative References . . . . . . . . . . . . . . . . . . . 8 66 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 8 68 1. Introduction 70 Operators of popular Internet services face complex challenges 71 associated with scaling their infrastructure. One approach is to 72 utilize equal-cost multi-path (ECMP) routing to perform stateless 73 distribution of incoming TCP or UDP sessions to multiple servers or 74 to middle boxes such as load balancers. Distribution of traffic in 75 this manner presents a problem when dealing with ICMP signaling. 76 Specifically, an ICMP error is not guaranteed to hash via ECMP to the 77 same destination as its corresponding TCP or UDP session. A case 78 where this is particularly problematic operationally is path MTU 79 discovery (PMTUD). 81 2. Problem 83 A common application for stateless load balancing of TCP or UDP flows 84 is to perform an initial subdivision of flows in front of a stateful 85 load balancer tier or multiple servers so that the workload becomes 86 divided into manageable fractions of the total number of flows. The 87 flow division is performed using ECMP forwarding and a stateless but 88 sticky algorithm for hashing across the available paths. This 89 nexthop selection for the purposes of flow distribution is a 90 constrained form of anycast topology where all anycast destinations 91 are equidistant from the upstream router responsible for making the 92 last next-hop forwarding decision before the flow arrives on the 93 destination device. In this approach, the hash is performed across 94 some set of available protocol headers. Typically, these headers may 95 include all or a subset of (IPv6) Flow-Label, IP-source, IP- 96 destination, protocol, source-port, destination-port and potentially 97 others such as ingress interface. 99 A problem common to this approach of distribution through hashing is 100 impact on path MTU discovery. An ICMPv6 type 2 PTB message generated 101 on an intermediate device for a packet sent from a server that is 102 part of an ECMP load balanced service to a client will have the load 103 balanced anycast address as the destination and hence will be 104 statelessly load balanced to one of the servers. While the ICMPv6 105 PTB message contains as much of the packet that could not be 106 forwarded as possible, the payload headers are not considered in the 107 forwarding decision and are ignored. Because the PTB message is not 108 identifiable as part of the original flow by the IP or upper layer 109 packet headers, the results of the ICMPv6 ECMP hash are unlikely to 110 be hashed to the same nexthop as packets matching TCP or UDP ECMP 111 hash. 113 An example packet flow and topology follow. 115 ptb -> router ecmp -> nexthop L4/L7 load balancer -> destination 117 router --> load balancer 1 ---> 118 \\--> load balancer 2 ---> load-balanced service 119 \--> load balancer N ---> 121 Figure 1 123 The router ECMP decision is used because it is part of the forwarding 124 architecture, can be performed at line rate, and does not depend on 125 shared state or coordination across a distributed forwarding system 126 which may include multiple linecards or routers. The ECMP routing 127 decision is deterministic with respect to packets having the same 128 computed hash. 130 A typical case where ICMPv6 PTB messages are received at the load 131 balancer is a case where the path MTU from the client to the load 132 balancer is limited by a tunnel in which the client itself is not 133 aware of. 135 Direct experience says that the frequency of PTB messages is small 136 compared to total flows. One possible conclusion being that tunneled 137 IPv6 deployments that cannot carry 1500 MTU packets are relatively 138 rare. Techniques employed by clients such as happy-eyeballs may 139 actually contribute some amelioration to the IPv6 client experience 140 by preferring IPv4 in cases that might be identified as failures. 142 Still, the expectation of operators is that PMTUD should work and 143 that unnecessary breakage of client traffic should be avoided. 145 A final observation regarding server tuning is that it is not always 146 possible even if it is potentially desirable to be able to 147 independently set the TCP MSS for different address families on some 148 end-systems. On Linux platforms, advmss may be set on a per route 149 basis for selected destinations in cases where discrimination by 150 route is possible. 152 The problem as described does also impact IPv4; however 153 implementation of RFC 4821 [RFC4821] TCP MTU probing, the ability to 154 fragment on wire at tunnel ingress points and the relative rarity of 155 sub-1500 byte MTUs that are not coupled to changes in client behavior 156 (for example, endpoint VPN clients set the tunnel interface MTU 157 accordingly for performance reasons) makes the problem sufficiently 158 rare that some existing deployments have choosen to ignore it. 160 3. Mitigation 162 Mitigation of the potential for PTB messages to be mis-delivered 163 involves ensuring that an ICMPv6 error message is distributed to the 164 same anycast server responsible for the flow for which the error is 165 generated. Ideally, mitigation could be done by the mechanism hosts 166 use to identify the flow, by looking into the payload of the ICMPv6 167 message (to determine which TCP flow it was associated with) before 168 making a forwarding decision. Because the encapsulated IP header 169 occurs at a fixed offset in the ICMP message it is not outside the 170 realm of possibility that routers with sufficient header processing 171 capability could parse that far into the payload. Employing a 172 mediation device that handles the parsing and distribution of PTB 173 messages after policy routing or on each load-balancer/server is a 174 possibility. 176 Another mitigation approach is predicated upon distributing the PTB 177 message to all anycast servers under the assumption that the one for 178 which the message was intended will be able to match it to the flow 179 and update the route cache with the new MTU and that devices not able 180 to match the flow will discard these packets. Such distribution has 181 potentially significant implications for resource consumption and the 182 potential for self-inflicted denial-of-service if not carefully 183 employed. Fortunately, in real-world deployments we have observed 184 that the number of flows for which this problem occurs is relatively 185 small (example, 10 or fewer pps on 1Gb/s or more worth of https 186 traffic) and sensible ingress rate limiters which will discard 187 excessive message volume can be applied to protect even very large 188 anycast server tiers with the potential for fallout only under 189 circumstances of deliberate duress. 191 3.1. Alternatives 193 As an alternative, it may be appropriate to lower the TCP MSS to 1220 194 in order to accommodate 1280 byte MTU. We consider this undesirable 195 as hosts may not be able to independently set TCP MSS by address- 196 family thereby impacting IPv4, or alternatively that middle-boxes 197 need to be employed to clamp the MSS independently from the end- 198 systems. Potentialy, extension might further alter the lower bound 199 that the mss would have to be set to making clamping still more 200 undesirable. 202 3.2. Implementation 204 1. Filter-based-forwarding matches next-header ICMPv6 type-2 and 205 matches a next-hop on a particular subnet directly attached to 206 both border routers. The filter is policed to reasonable limits 207 (we chose 1000pps more conservative rates might be required in 208 other imlementations). 210 2. Filter is applied on input side of all external interfaces 212 3. A proxy located at the next-hop forwards ICMPv6 type-2 packets 213 received at the next-hop to an Ethernet broadcast address 214 (example ff:ff:ff:ff:ff:ff) on all specified subnets. This was 215 necessitated by router inability (in IPv6) to forward the same 216 packet to multiple unicast next-hops. 218 4. Anycast servers receive the PTB error and process packet as 219 needed. 221 A simple Python scapy script that can perform the ICMPv6 proxy 222 reflection is included. 224 #!/usr/bin/python 226 from scapy.all import * 228 IFACE_OUT = ["p2p1", "p2p2"] 230 def icmp6_callback(pkt): 231 if pkt.haslayer(IPv6) and (ICMPv6PacketTooBig in pkt) \ 232 and pkt[Ether].dst != 'ff:ff:ff:ff:ff:ff': 233 del(pkt[Ether].src) 234 pkt[Ether].dst = 'ff:ff:ff:ff:ff:ff' 235 pkt.show() 236 for iface in IFACE_OUT: 237 sendp(pkt, iface=iface) 239 def main(): 240 sniff(prn=icmp6_callback, filter="icmp6 \ 241 and (ip6[40+0] == 2)", store=0) 243 if __name__ == '__main__': 244 main() 246 This example script listens on all interfaces for IPv6 PTB errors 247 being forwarded using filter-based-forwarding. It removes the 248 existing Ethernet source and rewrites a new Ethernet destination of 249 the Ethernet broadcast address. It then sends the resulting frame 250 out the p2p1 and p2p2 interfaces where our anycast servers reside. 252 3.2.1. Alternatives 254 Alternatively, network designs in which a common layer 2 network 255 exists on the ECMP hop could distribute the proxy onto the end 256 systems, eleminating the need for policy routing. They could then 257 rewrite the destination -- for example, using iptables before 258 forwarding the packet back to the network containing all of the 259 server or load balancer interfaces. This implmentation can be done 260 entirely within the Linux iptables firewall. Because of the 261 distributed nature of the filter, more conservative rate limits are 262 required than when a global rate limit can be employed. 264 An example ip6tables / nftables rule to match icmp6 traffic, not 265 match broadcast traffic, impose a rate limit of 10 pps, and pass to a 266 target destination would resemble: 268 ip6tables -I INPUT -i lo -p icmpv6 -m icmpv6 --icmpv6-type 2/0 \ 269 -m pkttype ! --pkt-type broadcast -m limit --limit 10/second \ 270 -j TEE 2001:DB8::1 272 As with the scapy example, once the destination has been rewritten 273 from a hardcoded ND entry to an Ethernet broadcast address -- in this 274 case to an IPv6 documentation address -- the traffic will be 275 reflected to all the hosts on the subnet. 277 4. Improvements 279 There are several ways that improvements could be made to the 280 situation with respect to ECMP load balancing of ICMPv6 PTB. 282 1. Routers with sufficient capacity within the lookup process could 283 parse all the way through the L3 or L4 header in the ICMPv6 284 payload beginning at bit offset 32 of the ICMP header. By 285 reordering the elements of the hash to match the inward direction 286 of the flow, the PTB error could be directed to the same next-hop 287 as the incoming packets in the flow. 289 2. The FIB (Forwarding Information Base) on the router could be 290 programmed with a multicast distribution tree that included all 291 of the necessary next-hops. 293 3. Ubiquitous implementation of RFC 4821 [RFC4821] Packetization 294 Layer Path MTU Discovery would probably go a long way towards 295 reducing dependence on ICMPv6 PTB. 297 5. Acknowledgements 299 The authors would like to thank Marak Majkowsiki for contributing 300 text, examples, and a very close review. The authors would like to 301 thank Mark Andrews, Brian Carpenter, Nick Hilliard and Ray Hunter, 302 for review. 304 6. IANA Considerations 306 This memo includes no request to IANA. 308 7. Security Considerations 310 The employed mitigation has the potential to greatly amplify the 311 impact of a deliberately malicious sending of ICMPv6 PTB messages. 312 Sensible ingress rate limiting can reduce the potential for impact; 313 however, legitimate traffic may be lost once the rate limit is 314 reached. 316 The proxy replication results in devices not associated with the flow 317 that generated the PTB being recipients of an ICMPv6 message which 318 contains a fragment of a packet. This could arguably result in 319 information disclosure. Recipient machines should be in a common 320 administrative domain. 322 8. Informative References 324 [RFC4821] Mathis, M. and J. Heffner, "Packetization Layer Path MTU 325 Discovery", RFC 4821, March 2007. 327 Authors' Addresses 329 Matt Byerly 330 Fastly 331 Kapolei, HI 332 US 334 Email: suckawha@gmail.com 336 Matt Hite 337 Evernote 338 Redwood City, CA 339 US 341 Email: mhite@hotmail.com 343 Joel Jaeggli 344 Fastly 345 Mountain View, CA 346 US 348 Email: joelja@gmail.com