| 1 |
<pre>Internet Engineering Task Force (IETF) S. Bryant, Ed. |
| 2 |
Request for Comments: 6391 C. Filsfils |
| 3 |
Category: Standards Track Cisco Systems |
| 4 |
ISSN: 2070-1721 U. Drafz |
| 5 |
Deutsche Telekom |
| 6 |
V. Kompella |
| 7 |
J. Regan |
| 8 |
Alcatel-Lucent |
| 9 |
S. Amante |
| 10 |
Level 3 Communications, LLC |
| 11 |
November 2011 |
| 12 |
|
| 13 |
|
| 14 |
Flow-Aware Transport of Pseudowires over an MPLS Packet Switched Network |
| 15 |
|
| 16 |
Abstract |
| 17 |
|
| 18 |
Where the payload of a pseudowire comprises a number of distinct |
| 19 |
flows, it can be desirable to carry those flows over the Equal Cost |
| 20 |
Multiple Paths (ECMPs) that exist in the packet switched network. |
| 21 |
Most forwarding engines are able to generate a hash of the MPLS label |
| 22 |
stack and use this mechanism to balance MPLS flows over ECMPs. |
| 23 |
|
| 24 |
This document describes a method of identifying the flows, or flow |
| 25 |
groups, within pseudowires such that Label Switching Routers can |
| 26 |
balance flows at a finer granularity than individual pseudowires. |
| 27 |
The mechanism uses an additional label in the MPLS label stack. |
| 28 |
|
| 29 |
Status of This Memo |
| 30 |
|
| 31 |
This is an Internet Standards Track document. |
| 32 |
|
| 33 |
This document is a product of the Internet Engineering Task Force |
| 34 |
(IETF). It represents the consensus of the IETF community. It has |
| 35 |
received public review and has been approved for publication by the |
| 36 |
Internet Engineering Steering Group (IESG). Further information on |
| 37 |
Internet Standards is available in <a href="./rfc5741#section-2">Section 2 of RFC 5741</a>. |
| 38 |
|
| 39 |
Information about the current status of this document, any errata, |
| 40 |
and how to provide feedback on it may be obtained at |
| 41 |
<a href="http://www.rfc-editor.org/info/rfc6391">http://www.rfc-editor.org/info/rfc6391</a>. |
| 42 |
|
| 43 |
|
| 44 |
|
| 45 |
|
| 46 |
|
| 47 |
|
| 48 |
|
| 49 |
|
| 50 |
|
| 51 |
|
| 52 |
<span class="grey">Bryant, et al. Standards Track [Page 1]</span></pre> |
| 53 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span> |
| 54 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 55 |
|
| 56 |
|
| 57 |
Copyright Notice |
| 58 |
|
| 59 |
Copyright (c) 2011 IETF Trust and the persons identified as the |
| 60 |
document authors. All rights reserved. |
| 61 |
|
| 62 |
This document is subject to <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a> and the IETF Trust's Legal |
| 63 |
Provisions Relating to IETF Documents |
| 64 |
(<a href="http://trustee.ietf.org/license-info">http://trustee.ietf.org/license-info</a>) in effect on the date of |
| 65 |
publication of this document. Please review these documents |
| 66 |
carefully, as they describe your rights and restrictions with respect |
| 67 |
to this document. Code Components extracted from this document must |
| 68 |
include Simplified BSD License text as described in Section 4.e of |
| 69 |
the Trust Legal Provisions and are provided without warranty as |
| 70 |
described in the Simplified BSD License. |
| 71 |
|
| 72 |
Table of Contents |
| 73 |
|
| 74 |
<a href="#section-1">1</a>. Introduction ....................................................<a href="#page-3">3</a> |
| 75 |
<a href="#section-1.1">1.1</a>. Requirements Language ......................................<a href="#page-4">4</a> |
| 76 |
<a href="#section-1.2">1.2</a>. ECMP in Label Switching Routers ............................<a href="#page-4">4</a> |
| 77 |
<a href="#section-1.3">1.3</a>. Flow Label .................................................<a href="#page-4">4</a> |
| 78 |
<a href="#section-2">2</a>. Native Service Processing Function ..............................<a href="#page-5">5</a> |
| 79 |
<a href="#section-3">3</a>. Pseudowire Forwarder ............................................<a href="#page-6">6</a> |
| 80 |
<a href="#section-3.1">3.1</a>. Encapsulation ..............................................<a href="#page-7">7</a> |
| 81 |
<a href="#section-4">4</a>. Signalling the Presence of the Flow Label .......................<a href="#page-8">8</a> |
| 82 |
<a href="#section-4.1">4.1</a>. Structure of Flow Label Sub-TLV ............................<a href="#page-9">9</a> |
| 83 |
<a href="#section-5">5</a>. Static Pseudowires ..............................................<a href="#page-9">9</a> |
| 84 |
<a href="#section-6">6</a>. Multi-Segment Pseudowires .......................................<a href="#page-9">9</a> |
| 85 |
<a href="#section-7">7</a>. Operations, Administration, and Maintenance (OAM) ..............<a href="#page-10">10</a> |
| 86 |
<a href="#section-8">8</a>. Applicability of PWs Using Flow Labels .........................<a href="#page-11">11</a> |
| 87 |
<a href="#section-8.1">8.1</a>. Equal Cost Multiple Paths .................................<a href="#page-12">12</a> |
| 88 |
<a href="#section-8.2">8.2</a>. Link Aggregation Groups ...................................<a href="#page-13">13</a> |
| 89 |
<a href="#section-8.3">8.3</a>. Multiple RSVP-TE Paths ....................................<a href="#page-13">13</a> |
| 90 |
<a href="#section-8.4">8.4</a>. The Single Large Flow Case ................................<a href="#page-14">14</a> |
| 91 |
<a href="#section-8.5">8.5</a>. Applicability to MPLS-TP ..................................<a href="#page-15">15</a> |
| 92 |
<a href="#section-8.6">8.6</a>. Asymmetric Operation ......................................<a href="#page-15">15</a> |
| 93 |
<a href="#section-9">9</a>. Applicability to MPLS LSPs .....................................<a href="#page-15">15</a> |
| 94 |
<a href="#section-10">10</a>. Security Considerations .......................................<a href="#page-16">16</a> |
| 95 |
<a href="#section-11">11</a>. IANA Considerations ...........................................<a href="#page-16">16</a> |
| 96 |
<a href="#section-12">12</a>. Congestion Considerations .....................................<a href="#page-16">16</a> |
| 97 |
<a href="#section-13">13</a>. Acknowledgements ..............................................<a href="#page-17">17</a> |
| 98 |
<a href="#section-14">14</a>. References ....................................................<a href="#page-17">17</a> |
| 99 |
<a href="#section-14.1">14.1</a>. Normative References .....................................<a href="#page-17">17</a> |
| 100 |
<a href="#section-14.2">14.2</a>. Informative References ...................................<a href="#page-18">18</a> |
| 101 |
|
| 102 |
|
| 103 |
|
| 104 |
|
| 105 |
|
| 106 |
|
| 107 |
|
| 108 |
<span class="grey">Bryant, et al. Standards Track [Page 2]</span></pre> |
| 109 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span> |
| 110 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 111 |
|
| 112 |
|
| 113 |
<span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span> |
| 114 |
|
| 115 |
A pseudowire (PW) [<a href="./rfc3985" title=""Pseudo Wire Emulation Edge-to-Edge (PWE3) Architecture"">RFC3985</a>] is normally transported over one single |
| 116 |
network path, even if multiple Equal Cost Multiple Paths (ECMPs) |
| 117 |
exist between the ingress and egress PW provider edge (PE) equipment |
| 118 |
[<a href="./rfc4385" title=""Pseudowire Emulation Edge-to-Edge (PWE3) Control Word for Use over an MPLS PSN"">RFC4385</a>] [<a href="./rfc4928" title=""Avoiding Equal Cost Multipath Treatment in MPLS Networks"">RFC4928</a>]. This is required to preserve the |
| 119 |
characteristics of the emulated service (e.g., to avoid misordering |
| 120 |
Structure-Agnostic Time Division Multiplexing over Packet (SAToP) PW |
| 121 |
packets [<a href="./rfc4553" title=""Structure- Agnostic Time Division Multiplexing (TDM) over Packet (SAToP)"">RFC4553</a>] or subjecting the packets to unusable inter-arrival |
| 122 |
times). The use of a single path to preserve order remains the |
| 123 |
default mode of operation of a PW. The new capability proposed in |
| 124 |
this document is an OPTIONAL mode that may be used when the use of |
| 125 |
ECMPs is known to be beneficial (and not harmful) to the operation of |
| 126 |
the PW. |
| 127 |
|
| 128 |
Some PWs are used to transport large volumes of IP traffic between |
| 129 |
routers. One example of this is the use of an Ethernet PW to create |
| 130 |
a virtual direct link between a pair of routers. Such PWs may carry |
| 131 |
from hundreds of Mbps to Gbps of traffic. These PWs only require |
| 132 |
packet ordering to be preserved within the context of each individual |
| 133 |
transported IP flow. They do not require packet ordering to be |
| 134 |
preserved between all packets of all IP flows within the pseudowire. |
| 135 |
|
| 136 |
The ability to explicitly configure such a PW to leverage the |
| 137 |
availability of multiple ECMPs allows for better capacity planning, |
| 138 |
as the statistical multiplexing of a larger number of smaller flows |
| 139 |
is more efficient than with a smaller set of larger flows. |
| 140 |
|
| 141 |
Typically, forwarding hardware can deduce that an IP payload is being |
| 142 |
directly carried by an MPLS label stack, and it is capable of looking |
| 143 |
at some fields in packets to construct hash buckets for conversations |
| 144 |
or flows. However, when the MPLS payload is a PW, an intermediate |
| 145 |
node has no information on the type of PW being carried in the |
| 146 |
packet. This limits the forwarder at the intermediate node to only |
| 147 |
being able to make an ECMP choice based on a hash of the MPLS label |
| 148 |
stack. In the case of a PW emulating a high-bandwidth trunk, the |
| 149 |
granularity obtained by hashing the label stack is inadequate for |
| 150 |
satisfactory load balancing. The ingress node, however, is in the |
| 151 |
special position of being able to understand the unencapsulated |
| 152 |
packet header to assist with spreading flows among any available |
| 153 |
ECMPs, or even any Loop-Free Alternates [<a href="./rfc5286" title=""Basic Specification for IP Fast Reroute: Loop-Free Alternates"">RFC5286</a>]. This document |
| 154 |
defines a method to introduce granularity on the hashing of traffic |
| 155 |
running over PWs by introducing an additional label, chosen by the |
| 156 |
ingress node, and placed at the bottom of the label stack. |
| 157 |
|
| 158 |
|
| 159 |
|
| 160 |
|
| 161 |
|
| 162 |
|
| 163 |
|
| 164 |
<span class="grey">Bryant, et al. Standards Track [Page 3]</span></pre> |
| 165 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span> |
| 166 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 167 |
|
| 168 |
|
| 169 |
In addition to providing an indication of the flow structure for use |
| 170 |
in ECMP forwarding decisions, the mechanism described in the document |
| 171 |
may also be used to select flows for distribution over an IEEE |
| 172 |
802.1AX-2008 (originally specified as IEEE 802.3ad-2000) Link |
| 173 |
Aggregation Group (LAG) that has been used in an MPLS network. |
| 174 |
|
| 175 |
NOTE: Although Ethernet is frequently referenced as a use case in |
| 176 |
this RFC, the mechanisms described in this document are general |
| 177 |
mechanisms that may be applied to any PW type in which there are |
| 178 |
identifiable flows, and in which there is no requirement to preserve |
| 179 |
the order between those flows. |
| 180 |
|
| 181 |
<span class="h3"><a class="selflink" id="section-1.1" href="#section-1.1">1.1</a>. Requirements Language</span> |
| 182 |
|
| 183 |
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", |
| 184 |
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this |
| 185 |
document are to be interpreted as described in <a href="./rfc2119">RFC 2119</a> [<a href="./rfc2119" title=""Key words for use in RFCs to Indicate Requirement Levels"">RFC2119</a>]. |
| 186 |
|
| 187 |
<span class="h3"><a class="selflink" id="section-1.2" href="#section-1.2">1.2</a>. ECMP in Label Switching Routers</span> |
| 188 |
|
| 189 |
Label Switching Routers (LSRs) commonly generate a hash of the label |
| 190 |
stack or some elements of the label stack as a method of |
| 191 |
discriminating between flows and use this to distribute those flows |
| 192 |
over the available ECMPs that exist in the network. Since the label |
| 193 |
at the bottom of the stack is usually the label most closely |
| 194 |
associated with the flow, this normally provides the greatest |
| 195 |
entropy, and hence is usually included in the hash. This document |
| 196 |
describes a method of adding an additional Label Stack Entry (LSE) at |
| 197 |
the bottom of the stack in order to facilitate the load balancing of |
| 198 |
the flows within a PW over the available ECMPs. A similar design for |
| 199 |
general MPLS use has also been proposed [<a href="#ref-MPLS-ENTROPY">MPLS-ENTROPY</a>]; see <a href="#section-9">Section 9</a> |
| 200 |
of this document. |
| 201 |
|
| 202 |
An alternative method of load balancing by creating a number of PWs |
| 203 |
and distributing the flows amongst them was considered, but was |
| 204 |
rejected because: |
| 205 |
|
| 206 |
o It did not introduce as much entropy as can be introduced by |
| 207 |
adding an additional LSE. |
| 208 |
|
| 209 |
o It required additional PWs to be set up and maintained. |
| 210 |
|
| 211 |
<span class="h3"><a class="selflink" id="section-1.3" href="#section-1.3">1.3</a>. Flow Label</span> |
| 212 |
|
| 213 |
An additional LSE [<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>] is interposed between the PW LSE and the |
| 214 |
control word, or if the control word is not present, between the PW |
| 215 |
LSE and the PW payload. This additional LSE is called the flow LSE, |
| 216 |
and the label carried by the flow LSE is called the flow label. |
| 217 |
|
| 218 |
|
| 219 |
|
| 220 |
<span class="grey">Bryant, et al. Standards Track [Page 4]</span></pre> |
| 221 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span> |
| 222 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 223 |
|
| 224 |
|
| 225 |
Indivisible flows within the PW MUST be mapped to the same flow label |
| 226 |
by the ingress PE. The flow label stimulates the correct ECMP load- |
| 227 |
balancing behaviour in the packet switched network (PSN). On receipt |
| 228 |
of the PW packet at the egress PE (which knows a flow LSE is |
| 229 |
present), the flow LSE is discarded without processing. |
| 230 |
|
| 231 |
Note that the flow label MUST NOT be an MPLS reserved label (values |
| 232 |
in the range 0..15) [<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>], but is otherwise unconstrained by the |
| 233 |
protocol. |
| 234 |
|
| 235 |
It is useful to give consideration to the choice of Time to Live |
| 236 |
(TTL) value in the flow LSE [<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>]. The flow LSE is at the bottom |
| 237 |
of the label stack; therefore, even when penultimate hop popping is |
| 238 |
employed, it will always be preceded by the PW label on arrival at |
| 239 |
the PE. If, due to an error condition, the flow LSE becomes the top |
| 240 |
of the stack, it might be examined as if it were a normal LSE, and |
| 241 |
the packet might then be forwarded. This can be prevented by setting |
| 242 |
the flow LSE TTL to 1, thereby forcing the packet to be discarded by |
| 243 |
the forwarder. Note that setting the TTL to 1 regardless of the |
| 244 |
payload may be considered a departure from the TTL procedures defined |
| 245 |
in [<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>] that apply to the general MPLS case. |
| 246 |
|
| 247 |
This document does not define a use for the Traffic Class (TC) field |
| 248 |
[<a href="./rfc5462" title=""Multiprotocol Label Switching (MPLS) Label Stack Entry: "">RFC5462</a>] (formerly known as the Experimental Use (EXP) bits |
| 249 |
[<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>]) in the flow label. Future documents may define a use for |
| 250 |
these bits; therefore, implementations conforming to this |
| 251 |
specification MUST set the TC field to zero at the ingress and MUST |
| 252 |
ignore them at the egress. |
| 253 |
|
| 254 |
<span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. Native Service Processing Function</span> |
| 255 |
|
| 256 |
The Native Service Processing (NSP) function [<a href="./rfc3985" title=""Pseudo Wire Emulation Edge-to-Edge (PWE3) Architecture"">RFC3985</a>] is a component |
| 257 |
of a PE that has knowledge of the structure of the emulated service |
| 258 |
and is able to take action on the service outside the scope of the |
| 259 |
PW. In this case, it is REQUIRED that the NSP in the ingress PE |
| 260 |
identify flows, or groups of flows within the service, and indicate |
| 261 |
the flow (group) identity of each packet as it is passed to the |
| 262 |
pseudowire forwarder. As an example, where the PW type is an |
| 263 |
Ethernet, the NSP might parse the ingress Ethernet traffic and |
| 264 |
consider all of the IP traffic. This traffic could then be |
| 265 |
categorised into flows by considering all traffic with the same |
| 266 |
source and destination address pair to be a single indivisible flow. |
| 267 |
Since this is an NSP function, by definition, the method used to |
| 268 |
identify a flow is outside the scope of the PW design. Similarly, |
| 269 |
since the NSP is internal to the PE, the method of flow indication to |
| 270 |
the PW forwarder is outside the scope of this document. |
| 271 |
|
| 272 |
|
| 273 |
|
| 274 |
|
| 275 |
|
| 276 |
<span class="grey">Bryant, et al. Standards Track [Page 5]</span></pre> |
| 277 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span> |
| 278 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 279 |
|
| 280 |
|
| 281 |
<span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Pseudowire Forwarder</span> |
| 282 |
|
| 283 |
The PW forwarder must be provided with a method of mapping flows to |
| 284 |
load-balanced paths. |
| 285 |
|
| 286 |
The forwarder must generate a label for the flow or group of flows. |
| 287 |
How the flow label values are determined is outside the scope of this |
| 288 |
document; however, the flow label allocated to a flow MUST NOT be an |
| 289 |
MPLS reserved label and SHOULD remain constant for the life of the |
| 290 |
flow. It is RECOMMENDED that the method chosen to generate the load- |
| 291 |
balancing labels introduce a high degree of entropy in their values, |
| 292 |
to maximise the entropy presented to the ECMP selection mechanism in |
| 293 |
the LSRs in the PSN, and hence distribute the flows as evenly as |
| 294 |
possible over the available PSN ECMP. The forwarder at the ingress |
| 295 |
PE prepends the PW control word (if applicable), and then pushes the |
| 296 |
flow label, followed by the PW label. |
| 297 |
|
| 298 |
NOTE: Although this document does not attempt to specify any hash |
| 299 |
algorithms, it is suggested that any such algorithm should be based |
| 300 |
on the assumption that there will be a high degree of entropy in the |
| 301 |
values assigned to the flow labels. |
| 302 |
|
| 303 |
The forwarder at the egress PE uses the pseudowire label to identify |
| 304 |
the pseudowire. From the context associated with the pseudowire |
| 305 |
label, the egress PE can determine whether a flow LSE is present. If |
| 306 |
a flow LSE is present, it MUST be checked to determine whether it |
| 307 |
carries a reserved label. If it is a reserved label, the packet is |
| 308 |
processed according to the rules associated with that reserved label; |
| 309 |
otherwise, the LSE is discarded. |
| 310 |
|
| 311 |
All other PW forwarding operations are unmodified by the inclusion of |
| 312 |
the flow LSE. |
| 313 |
|
| 314 |
|
| 315 |
|
| 316 |
|
| 317 |
|
| 318 |
|
| 319 |
|
| 320 |
|
| 321 |
|
| 322 |
|
| 323 |
|
| 324 |
|
| 325 |
|
| 326 |
|
| 327 |
|
| 328 |
|
| 329 |
|
| 330 |
|
| 331 |
|
| 332 |
<span class="grey">Bryant, et al. Standards Track [Page 6]</span></pre> |
| 333 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span> |
| 334 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 335 |
|
| 336 |
|
| 337 |
<span class="h3"><a class="selflink" id="section-3.1" href="#section-3.1">3.1</a>. Encapsulation</span> |
| 338 |
|
| 339 |
The PWE3 Protocol Stack Reference Model modified to include flow LSE |
| 340 |
is shown in Figure 1. |
| 341 |
|
| 342 |
+-------------+ +-------------+ |
| 343 |
| Emulated | | Emulated | |
| 344 |
| Ethernet | | Ethernet | |
| 345 |
| (including | Emulated Service | (including | |
| 346 |
| VLAN) |<==============================>| VLAN) | |
| 347 |
| Services | | Services | |
| 348 |
+-------------+ +-------------+ |
| 349 |
| Flow | | Flow | |
| 350 |
+-------------+ Pseudowire +-------------+ |
| 351 |
|Demultiplexer|<==============================>|Demultiplexer| |
| 352 |
+-------------+ +-------------+ |
| 353 |
| PSN | PSN Tunnel | PSN | |
| 354 |
| MPLS |<==============================>| MPLS | |
| 355 |
+-------------+ +-------------+ |
| 356 |
| Physical | | Physical | |
| 357 |
+-----+-------+ +-----+-------+ |
| 358 |
|
| 359 |
Figure 1: PWE3 Protocol Stack Reference Model |
| 360 |
|
| 361 |
The encapsulation of a PW with a flow LSE is shown in Figure 2. |
| 362 |
|
| 363 |
+---------------------------+ |
| 364 |
| | |
| 365 |
| Payload | |
| 366 |
| | n octets |
| 367 |
| | |
| 368 |
+---------------------------+ |
| 369 |
| Optional Control Word | 4 octets |
| 370 |
+---------------------------+ |
| 371 |
| Flow LSE | 4 octets |
| 372 |
+---------------------------+ |
| 373 |
| PW LSE | 4 octets |
| 374 |
+---------------------------+ |
| 375 |
| MPLS Tunnel LSE (s) | n*4 octets (four octets per LSE) |
| 376 |
+---------------------------+ |
| 377 |
|
| 378 |
Figure 2: Encapsulation of a Pseudowire with a Pseudowire Flow LSE |
| 379 |
|
| 380 |
|
| 381 |
|
| 382 |
|
| 383 |
|
| 384 |
|
| 385 |
|
| 386 |
|
| 387 |
|
| 388 |
<span class="grey">Bryant, et al. Standards Track [Page 7]</span></pre> |
| 389 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span> |
| 390 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 391 |
|
| 392 |
|
| 393 |
<span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Signalling the Presence of the Flow Label</span> |
| 394 |
|
| 395 |
When using the signalling procedures in [<a href="./rfc4447" title=""Pseudowire Setup and Maintenance Using the Label Distribution Protocol (LDP)"">RFC4447</a>], a new Pseudowire |
| 396 |
Interface Parameter Sub-TLV, the Flow Label Sub-TLV (FL Sub-TLV), is |
| 397 |
used to synchronise the flow label states between the ingress and |
| 398 |
egress PEs. |
| 399 |
|
| 400 |
The absence of an FL Sub-TLV indicates that the PE is unable to |
| 401 |
process flow labels. An ingress PE that is using PW signalling and |
| 402 |
that does not send an FL Sub-TLV MUST NOT include a flow label in the |
| 403 |
PW packet. An ingress PE that is using PW signalling and that does |
| 404 |
not receive an FL Sub-TLV from its egress peer MUST NOT include a |
| 405 |
flow label in the PW packet. This preserves backwards compatibility |
| 406 |
with existing PW specifications. |
| 407 |
|
| 408 |
A PE that wishes to send a flow label in a PW packet MUST include in |
| 409 |
its label mapping message an FL Sub-TLV with T = 1 (see <a href="#section-4.1">Section 4.1</a>). |
| 410 |
|
| 411 |
A PE that is willing to receive a flow label MUST include in its |
| 412 |
label mapping message an FL Sub-TLV with R = 1 (see <a href="#section-4.1">Section 4.1</a>). |
| 413 |
|
| 414 |
A PE that receives a label mapping message containing an FL Sub-TLV |
| 415 |
with R = 0 MUST NOT include a flow label in the PW packet. |
| 416 |
|
| 417 |
Thus, a PE sending an FL Sub-TLV with T = 1 and receiving an FL |
| 418 |
Sub-TLV with R = 1 MUST include a flow label in the PW packet. Under |
| 419 |
all other combinations of FL Sub-TLV signalling, a PE MUST NOT |
| 420 |
include a flow label in the PW packet. |
| 421 |
|
| 422 |
The signalling procedures in [<a href="./rfc4447" title=""Pseudowire Setup and Maintenance Using the Label Distribution Protocol (LDP)"">RFC4447</a>] state that "Processing of the |
| 423 |
interface parameters should continue when unknown interface |
| 424 |
parameters are encountered, and they MUST be silently ignored". The |
| 425 |
signalling procedure described here is therefore backwards compatible |
| 426 |
with existing implementations. |
| 427 |
|
| 428 |
Note that what is signalled is the desire to include the flow LSE in |
| 429 |
the label stack. The value of the flow label is a local matter for |
| 430 |
the ingress PE, and the label value itself is not signalled. |
| 431 |
|
| 432 |
|
| 433 |
|
| 434 |
|
| 435 |
|
| 436 |
|
| 437 |
|
| 438 |
|
| 439 |
|
| 440 |
|
| 441 |
|
| 442 |
|
| 443 |
|
| 444 |
<span class="grey">Bryant, et al. Standards Track [Page 8]</span></pre> |
| 445 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span> |
| 446 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 447 |
|
| 448 |
|
| 449 |
<span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. Structure of Flow Label Sub-TLV</span> |
| 450 |
|
| 451 |
The structure of the Flow Label Sub-TLV is shown in Figure 3. |
| 452 |
|
| 453 |
0 1 2 3 |
| 454 |
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 |
| 455 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 456 |
| FL=0x17 | Length |T|R| Reserved | |
| 457 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |
| 458 |
|
| 459 |
Figure 3: Flow Label Sub-TLV |
| 460 |
|
| 461 |
Where: |
| 462 |
|
| 463 |
o FL (value 0x17) is the Flow Label Sub-TLV identifier assigned by |
| 464 |
IANA (see <a href="#section-11">Section 11</a>). |
| 465 |
|
| 466 |
o Length is the length of the Sub-TLV in octets and is 4. |
| 467 |
|
| 468 |
o When T = 1, the PE is requesting the ability to send a PW packet |
| 469 |
that includes a flow label. When T = 0, the PE is indicating that |
| 470 |
it will not send a PW packet containing a flow label. |
| 471 |
|
| 472 |
o When R = 1, the PE is able to receive a PW packet with a flow |
| 473 |
label present. When R = 0, the PE is unable to receive a PW |
| 474 |
packet with the flow label present. |
| 475 |
|
| 476 |
o Reserved bits MUST be zero on transmit and MUST be ignored on |
| 477 |
receive. |
| 478 |
|
| 479 |
<span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Static Pseudowires</span> |
| 480 |
|
| 481 |
If PWE3 signalling [<a href="./rfc4447" title=""Pseudowire Setup and Maintenance Using the Label Distribution Protocol (LDP)"">RFC4447</a>] is not in use for a PW, then whether the |
| 482 |
flow label is used MUST be identically provisioned in both PEs at the |
| 483 |
PW endpoints. If there is no provisioning support for this option, |
| 484 |
the default behaviour is not to include the flow label. |
| 485 |
|
| 486 |
<span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. Multi-Segment Pseudowires</span> |
| 487 |
|
| 488 |
The flow label mechanism described in this document works on |
| 489 |
multi-segment PWs without requiring modification to the Switching PEs |
| 490 |
(S-PEs). This is because the flow LSE is transparent to the label |
| 491 |
swap operation, and because interface parameter Sub-TLV signalling is |
| 492 |
transitive. |
| 493 |
|
| 494 |
|
| 495 |
|
| 496 |
|
| 497 |
|
| 498 |
|
| 499 |
|
| 500 |
<span class="grey">Bryant, et al. Standards Track [Page 9]</span></pre> |
| 501 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-10" ></span> |
| 502 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 503 |
|
| 504 |
|
| 505 |
<span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. Operations, Administration, and Maintenance (OAM)</span> |
| 506 |
|
| 507 |
The following OAM considerations apply to this method of load |
| 508 |
balancing. |
| 509 |
|
| 510 |
Where the OAM is only to be used to perform a basic test to verify |
| 511 |
that the PWs have been configured at the PEs, Virtual Circuit |
| 512 |
Connectivity Verification (VCCV) [<a href="./rfc5085" title=""Pseudowire Virtual Circuit Connectivity Verification (VCCV): A Control Channel for Pseudowires"">RFC5085</a>] messages may be sent using |
| 513 |
any load balance PW path, i.e., using any value for the flow label. |
| 514 |
|
| 515 |
Where it is required to verify that a pseudowire is fully functional |
| 516 |
for all flows, a VCCV [<a href="./rfc5085" title=""Pseudowire Virtual Circuit Connectivity Verification (VCCV): A Control Channel for Pseudowires"">RFC5085</a>] connectivity verification message |
| 517 |
MUST be sent over each ECMP path to the pseudowire egress PE. This |
| 518 |
solution may be difficult to achieve and scales poorly. Under these |
| 519 |
circumstances, it may be sufficient to send VCCV messages using any |
| 520 |
load balance pseudowire path, because if a failure occurs within the |
| 521 |
PSN, the failure will normally be detected and repaired by the PSN. |
| 522 |
That is, the PSN's Interior Gateway Protocol (IGP) link/node failure |
| 523 |
detection mechanism (loss of light, bidirectional forwarding |
| 524 |
detection [<a href="./rfc5880" title=""Bidirectional Forwarding Detection (BFD)"">RFC5880</a>], or IGP hello detection) and the IGP convergence |
| 525 |
will naturally modify the ECMP set of network paths between the |
| 526 |
ingress and egress PEs. Hence, the PW is only impacted during the |
| 527 |
normal IGP convergence time. Note that this period may be reduced if |
| 528 |
a fast re-route or fast convergence technology is deployed in the |
| 529 |
network [<a href="./rfc4090" title=""Fast Reroute Extensions to RSVP-TE for LSP Tunnels"">RFC4090</a>] [<a href="./rfc5286" title=""Basic Specification for IP Fast Reroute: Loop-Free Alternates"">RFC5286</a>]. |
| 530 |
|
| 531 |
If the failure is related to the individual corruption of a Label |
| 532 |
Forwarding Information Base (LFIB) entry in a router, then only the |
| 533 |
network path using that specific entry is impacted. If the PW is |
| 534 |
load-balanced over multiple network paths, then this failure can only |
| 535 |
be detected if, by chance, the transported OAM flow is mapped onto |
| 536 |
the impacted network path, or if all paths are tested. Since testing |
| 537 |
all paths may present problems as noted above, other mechanisms to |
| 538 |
detect this type of error may need to be developed, such as a Label |
| 539 |
Switched Path (LSP) self-test technology. |
| 540 |
|
| 541 |
To troubleshoot the MPLS PSN, including multiple paths, the |
| 542 |
techniques described in [<a href="./rfc4378" title=""A Framework for Multi-Protocol Label Switching (MPLS) Operations and Management (OAM)"">RFC4378</a>] and [<a href="./rfc4379" title=""Detecting Multi-Protocol Label Switched (MPLS) Data Plane Failures"">RFC4379</a>] can be used. |
| 543 |
|
| 544 |
Where the PW OAM is carried out of band (VCCV Type 2) [<a href="./rfc5085" title=""Pseudowire Virtual Circuit Connectivity Verification (VCCV): A Control Channel for Pseudowires"">RFC5085</a>], it |
| 545 |
is necessary to insert an "MPLS Router Alert Label" in the label |
| 546 |
stack. The resultant label stack is as follows: |
| 547 |
|
| 548 |
|
| 549 |
|
| 550 |
|
| 551 |
|
| 552 |
|
| 553 |
|
| 554 |
|
| 555 |
|
| 556 |
<span class="grey">Bryant, et al. Standards Track [Page 10]</span></pre> |
| 557 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-11" ></span> |
| 558 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 559 |
|
| 560 |
|
| 561 |
+-------------------------------+ |
| 562 |
| | |
| 563 |
| VCCV Message | n octets |
| 564 |
| | |
| 565 |
+-------------------------------+ |
| 566 |
| Optional Control Word | 4 octets |
| 567 |
+-------------------------------+ |
| 568 |
| Flow LSE | 4 octets |
| 569 |
+-------------------------------+ |
| 570 |
| PW LSE | 4 octets |
| 571 |
+-------------------------------+ |
| 572 |
| Router Alert LSE | 4 octets |
| 573 |
+-------------------------------+ |
| 574 |
| MPLS Tunnel LSE(s) | n*4 octets (four octets per label) |
| 575 |
+-------------------------------+ |
| 576 |
|
| 577 |
Figure 4: Use of Router Alert Label |
| 578 |
|
| 579 |
Note that, depending on the number of labels hashed by the LSR, the |
| 580 |
inclusion of the Router Alert label may cause the OAM packet to be |
| 581 |
load-balanced to a different path from that taken by the data packets |
| 582 |
with identical flow and PW labels. |
| 583 |
|
| 584 |
<span class="h2"><a class="selflink" id="section-8" href="#section-8">8</a>. Applicability of PWs Using Flow Labels</span> |
| 585 |
|
| 586 |
A node within the PSN is not able to perform deep packet inspection |
| 587 |
(DPI) of the PW, as the PW technology is not self-describing: the |
| 588 |
structure of the PW payload is only known to the ingress and egress |
| 589 |
PE devices. The method proposed in this document provides a |
| 590 |
statistical mitigation of the problem of load balance in those cases |
| 591 |
where a PE is able to discern flows embedded in the traffic received |
| 592 |
on the attachment circuit. |
| 593 |
|
| 594 |
The methods described in this document are transparent to the PSN and |
| 595 |
as such do not require any new capability from the PSN. |
| 596 |
|
| 597 |
The requirement to load-balance over multiple PSN paths occurs when |
| 598 |
the ratio between the PW access speed and the PSN's core link |
| 599 |
bandwidth is large (e.g., >= 10%). ATM and Frame Relay are unlikely |
| 600 |
to meet this property. Ethernet may have this property, and for that |
| 601 |
reason this document focuses on Ethernet. Applications for other |
| 602 |
high-access-bandwidth PWs may be defined in the future. |
| 603 |
|
| 604 |
This design applies to MPLS PWs where it is meaningful to |
| 605 |
de-construct the packets presented to the ingress PE into flows. The |
| 606 |
mechanism described in this document promotes the distribution of |
| 607 |
flows within the PW over different network paths. In turn, this |
| 608 |
means that whilst packets within a flow are delivered in order |
| 609 |
|
| 610 |
|
| 611 |
|
| 612 |
<span class="grey">Bryant, et al. Standards Track [Page 11]</span></pre> |
| 613 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-12" ></span> |
| 614 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 615 |
|
| 616 |
|
| 617 |
(subject to normal IP delivery perturbations due to topology |
| 618 |
variation), order is no longer maintained for all packets sent over |
| 619 |
the PW. It is not proposed to associate a different sequence number |
| 620 |
with each flow. If sequence number support is required, the flow |
| 621 |
label mechanism MUST NOT be used. |
| 622 |
|
| 623 |
Where it is known that the traffic carried by the Ethernet PW is IP, |
| 624 |
the flows can be identified and mapped to an ECMP. Such methods |
| 625 |
typically include hashing on the source and destination addresses, |
| 626 |
the protocol ID and higher-layer flow-dependent fields such as |
| 627 |
TCP/UDP ports, Layer 2 Tunneling Protocol version 3 (L2TPv3) Session |
| 628 |
IDs, etc. |
| 629 |
|
| 630 |
Where it is known that the traffic carried by the Ethernet PW is |
| 631 |
non-IP, techniques used for link bundling between Ethernet switches |
| 632 |
may be reused. In this case, however, the latency distribution would |
| 633 |
be larger than is found in the link bundle case. The acceptability |
| 634 |
of the increased latency is for further study. Of particular |
| 635 |
importance, the Ethernet control frames SHOULD always be mapped to |
| 636 |
the same PSN path to ensure in-order delivery. |
| 637 |
|
| 638 |
<span class="h3"><a class="selflink" id="section-8.1" href="#section-8.1">8.1</a>. Equal Cost Multiple Paths</span> |
| 639 |
|
| 640 |
ECMP in packet switched networks is statistical in nature. The |
| 641 |
mapping of flows to a particular path does not take into account the |
| 642 |
bandwidth of the flow being mapped or the current bandwidth usage of |
| 643 |
the members of the ECMP set. This simplification works well when the |
| 644 |
distribution of flows is evenly spread over the ECMP set and there |
| 645 |
are a large number of flows that have low bandwidth relative to the |
| 646 |
paths. The random allocation of a flow to a path provides a good |
| 647 |
approximation to an even spread of flows, provided that polarisation |
| 648 |
effects are avoided. The method defined in this document has the |
| 649 |
same statistical properties as an IP PSN. |
| 650 |
|
| 651 |
ECMP is a load-sharing mechanism that is based on sharing the load |
| 652 |
over a number of layer 3 paths through the PSN. Often, however, |
| 653 |
multiple links exist between a pair of LSRs that are considered by |
| 654 |
the IGP to be a single link. These are known as link bundles. The |
| 655 |
mechanism described in this document can also be used to distribute |
| 656 |
the flows within a PW over the members of the link bundle by using |
| 657 |
the flow label value to identify candidate flows. How that mapping |
| 658 |
takes place is outside the scope of this specification. Similar |
| 659 |
considerations apply to Link Aggregation Groups. |
| 660 |
|
| 661 |
There is no mechanism currently defined to indicate the bandwidths in |
| 662 |
use by specific flows using the fields of the MPLS shim header. |
| 663 |
Furthermore, since the semantics of the MPLS shim header are fully |
| 664 |
defined in [<a href="./rfc3032" title=""MPLS Label Stack Encoding"">RFC3032</a>] and [<a href="./rfc5462" title=""Multiprotocol Label Switching (MPLS) Label Stack Entry: "">RFC5462</a>], those fields cannot be assigned |
| 665 |
|
| 666 |
|
| 667 |
|
| 668 |
<span class="grey">Bryant, et al. Standards Track [Page 12]</span></pre> |
| 669 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-13" ></span> |
| 670 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 671 |
|
| 672 |
|
| 673 |
semantics to carry this information. This document does not define |
| 674 |
any semantic for use in the TTL or TC fields of the label entry that |
| 675 |
carries the flow label, but requires that the flow label itself be |
| 676 |
selected with a high degree of entropy suggesting that the label |
| 677 |
value should not be overloaded with additional meaning in any |
| 678 |
subsequent specification. |
| 679 |
|
| 680 |
A different type of load balancing is the desire to carry a PW over a |
| 681 |
set of PSN links in which the bandwidth of members of the link set is |
| 682 |
less than the bandwidth of the PW. Proposals to address this problem |
| 683 |
have been made in the past [<a href="#ref-PWBONDING" title=""PW Bonding"">PWBONDING</a>]. Such a mechanism can be |
| 684 |
considered complementary to this mechanism. |
| 685 |
|
| 686 |
<span class="h3"><a class="selflink" id="section-8.2" href="#section-8.2">8.2</a>. Link Aggregation Groups</span> |
| 687 |
|
| 688 |
A Link Aggregation Group (LAG) is used to bond together several |
| 689 |
physical circuits between two adjacent nodes so they appear to |
| 690 |
higher-layer protocols as a single, higher-bandwidth "virtual" pipe. |
| 691 |
These may coexist in various parts of a given network. An advantage |
| 692 |
of LAGs is that they reduce the number of routing and signalling |
| 693 |
protocol adjacencies between devices, reducing control plane |
| 694 |
processing overhead. As with ECMP, the key problem related to LAGs |
| 695 |
is that due to inefficiencies in LAG load-distribution algorithms, a |
| 696 |
particular component of a LAG may experience congestion. The |
| 697 |
mechanism proposed here may be able to assist in producing a more |
| 698 |
uniform flow distribution. |
| 699 |
|
| 700 |
The same considerations requiring a flow to go over a single member |
| 701 |
of an ECMP set apply to a member of a LAG. |
| 702 |
|
| 703 |
<span class="h3"><a class="selflink" id="section-8.3" href="#section-8.3">8.3</a>. Multiple RSVP-TE Paths</span> |
| 704 |
|
| 705 |
In some networks, it is desirable for a Label Edge Router (LER) to be |
| 706 |
able to load-balance a PW across multiple Resource Reservation |
| 707 |
Protocol - Traffic Engineering (RSVP-TE) tunnels. The flow label |
| 708 |
mechanism described in this document may be used to provide the LER |
| 709 |
with the required flow information and necessary entropy to provide |
| 710 |
this type of load balancing. An example of such a case is the use of |
| 711 |
the flow label mechanism in networks using a link bundle with the all |
| 712 |
ones component [<a href="./rfc4201" title=""Link Bundling in MPLS Traffic Engineering (TE)"">RFC4201</a>]. |
| 713 |
|
| 714 |
Methods by which the LER is configured to apply this type of ECMP are |
| 715 |
outside the scope of this document. |
| 716 |
|
| 717 |
|
| 718 |
|
| 719 |
|
| 720 |
|
| 721 |
|
| 722 |
|
| 723 |
|
| 724 |
<span class="grey">Bryant, et al. Standards Track [Page 13]</span></pre> |
| 725 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-14" ></span> |
| 726 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 727 |
|
| 728 |
|
| 729 |
<span class="h3"><a class="selflink" id="section-8.4" href="#section-8.4">8.4</a>. The Single Large Flow Case</span> |
| 730 |
|
| 731 |
Clearly, the operator should make sure that the service offered using |
| 732 |
PW technology and the method described in this document do not exceed |
| 733 |
the maximum planned link capacity, unless it can be guaranteed that |
| 734 |
they conform to the Internet traffic profile of a very large number |
| 735 |
of small flows. |
| 736 |
|
| 737 |
If the NSP cannot access sufficient information to distinguish flows, |
| 738 |
perhaps because the protocol stack required parsing further into the |
| 739 |
packet than it is able, then the functionality described in this |
| 740 |
document does not give any benefits. The most common case where a |
| 741 |
single flow dominates the traffic on a PW is when it is used to |
| 742 |
transport enterprise traffic. Enterprise traffic may well consist of |
| 743 |
a single, large TCP flow, or encrypted flows that cannot be handled |
| 744 |
by the methods described in this document. |
| 745 |
|
| 746 |
An operator has four options under these circumstances: |
| 747 |
|
| 748 |
1. The operator can choose to do nothing, and the system will work |
| 749 |
as it does without the flow label. |
| 750 |
|
| 751 |
2. The operator can make the customer aware that the service |
| 752 |
offering has a restriction on flow bandwidth and police flows to |
| 753 |
that restriction. This would allow customers offering multiple |
| 754 |
flows to use a larger fraction of their access bandwidth, whilst |
| 755 |
preventing a single flow from consuming a fraction of internal |
| 756 |
link bandwidth that the operator considered excessive. |
| 757 |
|
| 758 |
3. The operator could configure the ingress PE to assign a constant |
| 759 |
flow label to all high-bandwidth flows so that only one path was |
| 760 |
affected by these flows. |
| 761 |
|
| 762 |
4. The operator could configure the ingress PE to assign a random |
| 763 |
flow label to all high-bandwidth flows so as to minimise the |
| 764 |
disruption to the network at the cost of out-of-order traffic to |
| 765 |
the user. |
| 766 |
|
| 767 |
The issues described above are mitigated by the following two |
| 768 |
factors: |
| 769 |
|
| 770 |
o Firstly, the customer of a high-bandwidth PW service has an |
| 771 |
incentive to get the best transport service, because an |
| 772 |
inefficient use of the PSN leads to jitter and eventually to loss |
| 773 |
to the PW's payload. |
| 774 |
|
| 775 |
|
| 776 |
|
| 777 |
|
| 778 |
|
| 779 |
|
| 780 |
<span class="grey">Bryant, et al. Standards Track [Page 14]</span></pre> |
| 781 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-15" ></span> |
| 782 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 783 |
|
| 784 |
|
| 785 |
o Secondly, the customer is usually able to tailor their |
| 786 |
applications to generate many flows in the PSN. A well-known |
| 787 |
example is massive data transport between servers that use many |
| 788 |
parallel TCP sessions. This same technique can be used by any |
| 789 |
transport protocol: multiple UDP ports, multiple L2TPv3 Session |
| 790 |
IDs, or multiple Generic Routing Encapsulation (GRE) keys may be |
| 791 |
used to decompose a large flow into smaller components. This |
| 792 |
approach may be applied to IPsec [<a href="./rfc4301" title=""Security Architecture for the Internet Protocol"">RFC4301</a>] where multiple Security |
| 793 |
Parameter Indexes (SPIs) may be allocated to the same security |
| 794 |
association. |
| 795 |
|
| 796 |
<span class="h3"><a class="selflink" id="section-8.5" href="#section-8.5">8.5</a>. Applicability to MPLS-TP</span> |
| 797 |
|
| 798 |
The MPLS Transport Profile (MPLS-TP) [<a href="./rfc5654" title=""Requirements of an MPLS Transport Profile"">RFC5654</a>] Requirement 44 states |
| 799 |
that "MPLS-TP MUST support mechanisms that ensure the integrity of |
| 800 |
the transported customer's service traffic as required by its |
| 801 |
associated Service Level Agreement (SLA). Loss of integrity may be |
| 802 |
defined as packet corruption, reordering, or loss during normal |
| 803 |
network conditions". In addition, MPLS-TP makes extensive use of the |
| 804 |
fate sharing between OAM and data packets, which is defeated by the |
| 805 |
flow LSE. The flow-aware transport of a PW reorders packets and |
| 806 |
therefore MUST NOT be deployed in a network conforming to MPLS-TP, |
| 807 |
unless these integrity requirements specified in the SLA can be |
| 808 |
satisfied. |
| 809 |
|
| 810 |
<span class="h3"><a class="selflink" id="section-8.6" href="#section-8.6">8.6</a>. Asymmetric Operation</span> |
| 811 |
|
| 812 |
The protocol defined in this document supports the asymmetric |
| 813 |
inclusion of the flow LSE. Asymmetric operation can be expected when |
| 814 |
there is asymmetry in the bandwidth requirements making it |
| 815 |
unprofitable for one PE to perform the flow classification, or when |
| 816 |
that PE is otherwise unable to perform the classification but is able |
| 817 |
to receive flow labeled packets from its peer. Asymmetric operation |
| 818 |
of the PW may also be required when one PE has a high transmission |
| 819 |
bandwidth requirement, but has a need to receive the entire PW on a |
| 820 |
single interface in order to perform a processing operation that |
| 821 |
requires the context of the complete PW (for example, policing of the |
| 822 |
egress traffic). |
| 823 |
|
| 824 |
<span class="h2"><a class="selflink" id="section-9" href="#section-9">9</a>. Applicability to MPLS LSPs</span> |
| 825 |
|
| 826 |
An extension of this technique is to create a basis for hash |
| 827 |
diversity without having to peek below the label stack for IP traffic |
| 828 |
carried over Label Distribution Protocol (LDP) LSPs. The |
| 829 |
generalisation of this extension to MPLS has been described in |
| 830 |
[<a href="#ref-MPLS-ENTROPY">MPLS-ENTROPY</a>]. This generalisation can be regarded as a |
| 831 |
|
| 832 |
|
| 833 |
|
| 834 |
|
| 835 |
|
| 836 |
<span class="grey">Bryant, et al. Standards Track [Page 15]</span></pre> |
| 837 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-16" ></span> |
| 838 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 839 |
|
| 840 |
|
| 841 |
complementary, but distinct, approach from the technique described in |
| 842 |
this document. While similar consideration may apply to the |
| 843 |
identification of flows and the allocation of flow label values, the |
| 844 |
flow labels are imposed by different network components, and the |
| 845 |
associated signalling mechanisms are different. |
| 846 |
|
| 847 |
<span class="h2"><a class="selflink" id="section-10" href="#section-10">10</a>. Security Considerations</span> |
| 848 |
|
| 849 |
The PW generic security considerations described in [<a href="./rfc3985" title=""Pseudo Wire Emulation Edge-to-Edge (PWE3) Architecture"">RFC3985</a>] and the |
| 850 |
security considerations applicable to a specific PW type (for |
| 851 |
example, in the case of an Ethernet PW [<a href="./rfc4448" title=""Encapsulation Methods for Transport of Ethernet over MPLS Networks"">RFC4448</a>]) apply. The |
| 852 |
security considerations in [<a href="./rfc5920" title=""Security Framework for MPLS and GMPLS Networks"">RFC5920</a>] also apply. |
| 853 |
|
| 854 |
<a href="#section-1.3">Section 1.3</a> describes considerations that apply to the TTL value used |
| 855 |
in the flow LSE. The use of a TTL value of one prevents the |
| 856 |
accidental forwarding of a packet based on the label value in the |
| 857 |
flow LSE. |
| 858 |
|
| 859 |
<span class="h2"><a class="selflink" id="section-11" href="#section-11">11</a>. IANA Considerations</span> |
| 860 |
|
| 861 |
IANA maintains the registry "Pseudowire Name Spaces (PWE3)" with |
| 862 |
sub-registry "Pseudowire Interface Parameters Sub-TLV type Registry". |
| 863 |
IANA has registered the Flow Label Sub-TLV type in this registry. |
| 864 |
|
| 865 |
Parameter ID Length Description Reference |
| 866 |
------------------------------------------------------ |
| 867 |
0x17 4 Flow Label <a href="./rfc6391">RFC 6391</a> |
| 868 |
|
| 869 |
<span class="h2"><a class="selflink" id="section-12" href="#section-12">12</a>. Congestion Considerations</span> |
| 870 |
|
| 871 |
The congestion considerations applicable to PWs as described in |
| 872 |
[<a href="./rfc3985" title=""Pseudo Wire Emulation Edge-to-Edge (PWE3) Architecture"">RFC3985</a>] apply to this design. |
| 873 |
|
| 874 |
The ability to explicitly configure a PW to leverage the availability |
| 875 |
of multiple ECMPs is beneficial to capacity planning as, all other |
| 876 |
parameters being constant, the statistical multiplexing of a larger |
| 877 |
number of smaller flows is more efficient than with a smaller number |
| 878 |
of larger flows. |
| 879 |
|
| 880 |
Note that if the classification into flows is only performed on IP |
| 881 |
packets, the behaviour of those flows in the face of congestion will |
| 882 |
be as already defined by the IETF for packets of that type, and no |
| 883 |
additional congestion processing is required. |
| 884 |
|
| 885 |
Where flows that are not IP are classified, PW congestion avoidance |
| 886 |
must be applied to each non-IP load balance group. |
| 887 |
|
| 888 |
|
| 889 |
|
| 890 |
|
| 891 |
|
| 892 |
<span class="grey">Bryant, et al. Standards Track [Page 16]</span></pre> |
| 893 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-17" ></span> |
| 894 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 895 |
|
| 896 |
|
| 897 |
<span class="h2"><a class="selflink" id="section-13" href="#section-13">13</a>. Acknowledgements</span> |
| 898 |
|
| 899 |
The authors wish to thank Mary Barnes, Eric Grey, Kireeti Kompella, |
| 900 |
Joerg Kuechemann, Wilfried Maas, Luca Martini, Mark Townsley, Rolf |
| 901 |
Winter, and Lucy Yong for valuable comments on this document. |
| 902 |
|
| 903 |
<span class="h2"><a class="selflink" id="section-14" href="#section-14">14</a>. References</span> |
| 904 |
|
| 905 |
<span class="h3"><a class="selflink" id="section-14.1" href="#section-14.1">14.1</a>. Normative References</span> |
| 906 |
|
| 907 |
[<a id="ref-RFC2119">RFC2119</a>] Bradner, S., "Key words for use in RFCs to Indicate |
| 908 |
Requirement Levels", <a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a>, <a href="./rfc2119">RFC 2119</a>, March 1997. |
| 909 |
|
| 910 |
[<a id="ref-RFC3032">RFC3032</a>] Rosen, E., Tappan, D., Fedorkow, G., Rekhter, Y., |
| 911 |
Farinacci, D., Li, T., and A. Conta, "MPLS Label Stack |
| 912 |
Encoding", <a href="./rfc3032">RFC 3032</a>, January 2001. |
| 913 |
|
| 914 |
[<a id="ref-RFC4379">RFC4379</a>] Kompella, K. and G. Swallow, "Detecting Multi-Protocol |
| 915 |
Label Switched (MPLS) Data Plane Failures", <a href="./rfc4379">RFC 4379</a>, |
| 916 |
February 2006. |
| 917 |
|
| 918 |
[<a id="ref-RFC4385">RFC4385</a>] Bryant, S., Swallow, G., Martini, L., and D. McPherson, |
| 919 |
"Pseudowire Emulation Edge-to-Edge (PWE3) Control Word |
| 920 |
for Use over an MPLS PSN", <a href="./rfc4385">RFC 4385</a>, February 2006. |
| 921 |
|
| 922 |
[<a id="ref-RFC4447">RFC4447</a>] Martini, L., Ed., Rosen, E., El-Aawar, N., Smith, T., and |
| 923 |
G. Heron, "Pseudowire Setup and Maintenance Using the |
| 924 |
Label Distribution Protocol (LDP)", <a href="./rfc4447">RFC 4447</a>, April 2006. |
| 925 |
|
| 926 |
[<a id="ref-RFC4448">RFC4448</a>] Martini, L., Ed., Rosen, E., El-Aawar, N., and G. Heron, |
| 927 |
"Encapsulation Methods for Transport of Ethernet over |
| 928 |
MPLS Networks", <a href="./rfc4448">RFC 4448</a>, April 2006. |
| 929 |
|
| 930 |
[<a id="ref-RFC4553">RFC4553</a>] Vainshtein, A., Ed., and YJ. Stein, Ed., "Structure- |
| 931 |
Agnostic Time Division Multiplexing (TDM) over Packet |
| 932 |
(SAToP)", <a href="./rfc4553">RFC 4553</a>, June 2006. |
| 933 |
|
| 934 |
[<a id="ref-RFC4928">RFC4928</a>] Swallow, G., Bryant, S., and L. Andersson, "Avoiding |
| 935 |
Equal Cost Multipath Treatment in MPLS Networks", |
| 936 |
<a href="https://www.rfc-editor.org/bcp/bcp128">BCP 128</a>, <a href="./rfc4928">RFC 4928</a>, June 2007. |
| 937 |
|
| 938 |
[<a id="ref-RFC5085">RFC5085</a>] Nadeau, T., Ed., and C. Pignataro, Ed., "Pseudowire |
| 939 |
Virtual Circuit Connectivity Verification (VCCV): A |
| 940 |
Control Channel for Pseudowires", <a href="./rfc5085">RFC 5085</a>, |
| 941 |
December 2007. |
| 942 |
|
| 943 |
|
| 944 |
|
| 945 |
|
| 946 |
|
| 947 |
|
| 948 |
<span class="grey">Bryant, et al. Standards Track [Page 17]</span></pre> |
| 949 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-18" ></span> |
| 950 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 951 |
|
| 952 |
|
| 953 |
<span class="h3"><a class="selflink" id="section-14.2" href="#section-14.2">14.2</a>. Informative References</span> |
| 954 |
|
| 955 |
[<a id="ref-MPLS-ENTROPY">MPLS-ENTROPY</a>] |
| 956 |
Kompella, K., Drake, J., Amante, S., Henderickx, W., and |
| 957 |
L. Yong, "The Use of Entropy Labels in MPLS Forwarding", |
| 958 |
Work in Progress, October 2011. |
| 959 |
|
| 960 |
[<a id="ref-PWBONDING">PWBONDING</a>] Stein, Y(J)., Mendelsohn, I., and R. Insler, "PW |
| 961 |
Bonding", Work in Progress, November 2008. |
| 962 |
|
| 963 |
[<a id="ref-RFC3985">RFC3985</a>] Bryant, S., Ed., and P. Pate, Ed., "Pseudo Wire Emulation |
| 964 |
Edge-to-Edge (PWE3) Architecture", <a href="./rfc3985">RFC 3985</a>, March 2005. |
| 965 |
|
| 966 |
[<a id="ref-RFC4090">RFC4090</a>] Pan, P., Ed., Swallow, G., Ed., and A. Atlas, Ed., "Fast |
| 967 |
Reroute Extensions to RSVP-TE for LSP Tunnels", <a href="./rfc4090">RFC 4090</a>, |
| 968 |
May 2005. |
| 969 |
|
| 970 |
[<a id="ref-RFC4201">RFC4201</a>] Kompella, K., Rekhter, Y., and L. Berger, "Link Bundling |
| 971 |
in MPLS Traffic Engineering (TE)", <a href="./rfc4201">RFC 4201</a>, |
| 972 |
October 2005. |
| 973 |
|
| 974 |
[<a id="ref-RFC4301">RFC4301</a>] Kent, S. and K. Seo, "Security Architecture for the |
| 975 |
Internet Protocol", <a href="./rfc4301">RFC 4301</a>, December 2005. |
| 976 |
|
| 977 |
[<a id="ref-RFC4378">RFC4378</a>] Allan, D., Ed., and T. Nadeau, Ed., "A Framework for |
| 978 |
Multi-Protocol Label Switching (MPLS) Operations and |
| 979 |
Management (OAM)", <a href="./rfc4378">RFC 4378</a>, February 2006. |
| 980 |
|
| 981 |
[<a id="ref-RFC5286">RFC5286</a>] Atlas, A., Ed., and A. Zinin, Ed., "Basic Specification |
| 982 |
for IP Fast Reroute: Loop-Free Alternates", <a href="./rfc5286">RFC 5286</a>, |
| 983 |
September 2008. |
| 984 |
|
| 985 |
[<a id="ref-RFC5462">RFC5462</a>] Andersson, L. and R. Asati, "Multiprotocol Label |
| 986 |
Switching (MPLS) Label Stack Entry: "EXP" Field Renamed |
| 987 |
to "Traffic Class" Field", <a href="./rfc5462">RFC 5462</a>, February 2009. |
| 988 |
|
| 989 |
[<a id="ref-RFC5654">RFC5654</a>] Niven-Jenkins, B., Ed., Brungard, D., Ed., Betts, M., |
| 990 |
Ed., Sprecher, N., and S. Ueno, "Requirements of an MPLS |
| 991 |
Transport Profile", <a href="./rfc5654">RFC 5654</a>, September 2009. |
| 992 |
|
| 993 |
[<a id="ref-RFC5880">RFC5880</a>] Katz, D. and D. Ward, "Bidirectional Forwarding Detection |
| 994 |
(BFD)", <a href="./rfc5880">RFC 5880</a>, June 2010. |
| 995 |
|
| 996 |
[<a id="ref-RFC5920">RFC5920</a>] Fang, L., Ed., "Security Framework for MPLS and GMPLS |
| 997 |
Networks", <a href="./rfc5920">RFC 5920</a>, July 2010. |
| 998 |
|
| 999 |
|
| 1000 |
|
| 1001 |
|
| 1002 |
|
| 1003 |
|
| 1004 |
<span class="grey">Bryant, et al. Standards Track [Page 18]</span></pre> |
| 1005 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-19" ></span> |
| 1006 |
<span class="grey"><a href="./rfc6391">RFC 6391</a> FAT-PW November 2011</span> |
| 1007 |
|
| 1008 |
|
| 1009 |
Authors' Addresses |
| 1010 |
|
| 1011 |
Stewart Bryant (editor) |
| 1012 |
Cisco Systems |
| 1013 |
250 Longwater Ave. |
| 1014 |
Reading RG2 6GB |
| 1015 |
United Kingdom |
| 1016 |
|
| 1017 |
Phone: +44-208-824-8828 |
| 1018 |
EMail: [email protected] |
| 1019 |
|
| 1020 |
|
| 1021 |
Clarence Filsfils |
| 1022 |
Cisco Systems |
| 1023 |
Brussels |
| 1024 |
Belgium |
| 1025 |
|
| 1026 |
EMail: [email protected] |
| 1027 |
|
| 1028 |
|
| 1029 |
Ulrich Drafz |
| 1030 |
Deutsche Telekom |
| 1031 |
Muenster |
| 1032 |
Germany |
| 1033 |
|
| 1034 |
EMail: [email protected] |
| 1035 |
|
| 1036 |
|
| 1037 |
Vach Kompella |
| 1038 |
Alcatel-Lucent |
| 1039 |
|
| 1040 |
EMail: [email protected] |
| 1041 |
|
| 1042 |
|
| 1043 |
Joe Regan |
| 1044 |
Alcatel-Lucent |
| 1045 |
|
| 1046 |
EMail: [email protected] |
| 1047 |
|
| 1048 |
|
| 1049 |
Shane Amante |
| 1050 |
Level 3 Communications, LLC |
| 1051 |
1025 Eldorado Blvd. |
| 1052 |
Broomfield, CO 80021 |
| 1053 |
USA |
| 1054 |
|
| 1055 |
EMail: [email protected] |
| 1056 |
|
| 1057 |
|
| 1058 |
|
| 1059 |
|
| 1060 |
Bryant, et al. Standards Track [Page 19] |
| 1061 |
</pre> |