| 1 |
<pre>Internet Engineering Task Force (IETF) IJ. Wijnands, Ed. |
| 2 |
Request for Comments: 8279 Cisco Systems, Inc. |
| 3 |
Category: Experimental E. Rosen, Ed. |
| 4 |
ISSN: 2070-1721 Juniper Networks, Inc. |
| 5 |
A. Dolganow |
| 6 |
Nokia |
| 7 |
T. Przygienda |
| 8 |
Juniper Networks, Inc. |
| 9 |
S. Aldrin |
| 10 |
Google, Inc. |
| 11 |
November 2017 |
| 12 |
|
| 13 |
|
| 14 |
<span class="h1">Multicast Using Bit Index Explicit Replication (BIER)</span> |
| 15 |
|
| 16 |
Abstract |
| 17 |
|
| 18 |
This document specifies a new architecture for the forwarding of |
| 19 |
multicast data packets. It provides optimal forwarding of multicast |
| 20 |
packets through a "multicast domain". However, it does not require a |
| 21 |
protocol for explicitly building multicast distribution trees, nor |
| 22 |
does it require intermediate nodes to maintain any per-flow state. |
| 23 |
This architecture is known as "Bit Index Explicit Replication" |
| 24 |
(BIER). When a multicast data packet enters the domain, the ingress |
| 25 |
router determines the set of egress routers to which the packet needs |
| 26 |
to be sent. The ingress router then encapsulates the packet in a |
| 27 |
BIER header. The BIER header contains a bit string in which each bit |
| 28 |
represents exactly one egress router in the domain; to forward the |
| 29 |
packet to a given set of egress routers, the bits corresponding to |
| 30 |
those routers are set in the BIER header. The procedures for |
| 31 |
forwarding a packet based on its BIER header are specified in this |
| 32 |
document. Elimination of the per-flow state and the explicit tree- |
| 33 |
building protocols results in a considerable simplification. |
| 34 |
|
| 35 |
|
| 36 |
|
| 37 |
|
| 38 |
|
| 39 |
|
| 40 |
|
| 41 |
|
| 42 |
|
| 43 |
|
| 44 |
|
| 45 |
|
| 46 |
|
| 47 |
|
| 48 |
|
| 49 |
|
| 50 |
|
| 51 |
|
| 52 |
<span class="grey">Wijnands, et al. Experimental [Page 1]</span></pre> |
| 53 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span> |
| 54 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 55 |
|
| 56 |
|
| 57 |
Status of This Memo |
| 58 |
|
| 59 |
This document is not an Internet Standards Track specification; it is |
| 60 |
published for examination, experimental implementation, and |
| 61 |
evaluation. |
| 62 |
|
| 63 |
This document defines an Experimental Protocol for the Internet |
| 64 |
community. This document is a product of the Internet Engineering |
| 65 |
Task Force (IETF). It represents the consensus of the IETF |
| 66 |
community. It has received public review and has been approved for |
| 67 |
publication by the Internet Engineering Steering Group (IESG). Not |
| 68 |
all documents approved by the IESG are a candidate for any level of |
| 69 |
Internet Standard; see <a href="./rfc7841#section-2">Section 2 of RFC 7841</a>. |
| 70 |
|
| 71 |
Information about the current status of this document, any errata, |
| 72 |
and how to provide feedback on it may be obtained at |
| 73 |
<a href="https://www.rfc-editor.org/info/rfc8279">https://www.rfc-editor.org/info/rfc8279</a>. |
| 74 |
|
| 75 |
Copyright Notice |
| 76 |
|
| 77 |
Copyright (c) 2017 IETF Trust and the persons identified as the |
| 78 |
document authors. All rights reserved. |
| 79 |
|
| 80 |
This document is subject to <a href="https://www.rfc-editor.org/bcp/bcp78">BCP 78</a> and the IETF Trust's Legal |
| 81 |
Provisions Relating to IETF Documents |
| 82 |
(<a href="https://trustee.ietf.org/license-info">https://trustee.ietf.org/license-info</a>) in effect on the date of |
| 83 |
publication of this document. Please review these documents |
| 84 |
carefully, as they describe your rights and restrictions with respect |
| 85 |
to this document. Code Components extracted from this document must |
| 86 |
include Simplified BSD License text as described in Section 4.e of |
| 87 |
the Trust Legal Provisions and are provided without warranty as |
| 88 |
described in the Simplified BSD License. |
| 89 |
|
| 90 |
|
| 91 |
|
| 92 |
|
| 93 |
|
| 94 |
|
| 95 |
|
| 96 |
|
| 97 |
|
| 98 |
|
| 99 |
|
| 100 |
|
| 101 |
|
| 102 |
|
| 103 |
|
| 104 |
|
| 105 |
|
| 106 |
|
| 107 |
|
| 108 |
<span class="grey">Wijnands, et al. Experimental [Page 2]</span></pre> |
| 109 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span> |
| 110 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 111 |
|
| 112 |
|
| 113 |
Table of Contents |
| 114 |
|
| 115 |
<a href="#section-1">1</a>. Introduction ....................................................<a href="#page-4">4</a> |
| 116 |
<a href="#section-2">2</a>. The BFR Identifier and BFR-Prefix ...............................<a href="#page-7">7</a> |
| 117 |
<a href="#section-3">3</a>. Encoding BFR Identifiers in BitStrings ..........................<a href="#page-8">8</a> |
| 118 |
<a href="#section-4">4</a>. Layering .......................................................<a href="#page-11">11</a> |
| 119 |
<a href="#section-4.1">4.1</a>. The Routing Underlay ......................................<a href="#page-11">11</a> |
| 120 |
<a href="#section-4.2">4.2</a>. The BIER Layer ............................................<a href="#page-12">12</a> |
| 121 |
<a href="#section-4.3">4.3</a>. The Multicast Flow Overlay ................................<a href="#page-13">13</a> |
| 122 |
<a href="#section-5">5</a>. Advertising BFR-ids and BFR-Prefixes ...........................<a href="#page-13">13</a> |
| 123 |
<a href="#section-6">6</a>. BIER Intra-Domain Forwarding Procedures ........................<a href="#page-15">15</a> |
| 124 |
<a href="#section-6.1">6.1</a>. Overview ..................................................<a href="#page-15">15</a> |
| 125 |
<a href="#section-6.2">6.2</a>. BFR Neighbors .............................................<a href="#page-17">17</a> |
| 126 |
<a href="#section-6.3">6.3</a>. The Bit Index Routing Table ...............................<a href="#page-18">18</a> |
| 127 |
<a href="#section-6.4">6.4</a>. The Bit Index Forwarding Table ............................<a href="#page-19">19</a> |
| 128 |
<a href="#section-6.5">6.5</a>. The BIER Forwarding Procedure .............................<a href="#page-20">20</a> |
| 129 |
<a href="#section-6.6">6.6</a>. Examples of BIER Forwarding ...............................<a href="#page-23">23</a> |
| 130 |
<a href="#section-6.6.1">6.6.1</a>. Example 1 ..........................................<a href="#page-23">23</a> |
| 131 |
<a href="#section-6.6.2">6.6.2</a>. Example 2 ..........................................<a href="#page-24">24</a> |
| 132 |
<a href="#section-6.7">6.7</a>. Equal-Cost Multipath Forwarding ...........................<a href="#page-26">26</a> |
| 133 |
<a href="#section-6.7.1">6.7.1</a>. Non-deterministic ECMP .............................<a href="#page-27">27</a> |
| 134 |
<a href="#section-6.7.2">6.7.2</a>. Deterministic ECMP .................................<a href="#page-28">28</a> |
| 135 |
<a href="#section-6.8">6.8</a>. Prevention of Loops and Duplicates ........................<a href="#page-29">29</a> |
| 136 |
<a href="#section-6.9">6.9</a>. When Some Nodes Do Not Support BIER .......................<a href="#page-30">30</a> |
| 137 |
<a href="#section-6.10">6.10</a>. Use of Different BitStringLengths within a Domain ........<a href="#page-33">33</a> |
| 138 |
<a href="#section-6.10.1">6.10.1</a>. BitStringLength Compatibility Check ...............<a href="#page-34">34</a> |
| 139 |
<a href="#section-6.10.2">6.10.2</a>. Handling BitStringLength Mismatches ...............<a href="#page-36">36</a> |
| 140 |
6.10.3. Transitioning from One BitStringLength to |
| 141 |
Another ...........................................<a href="#page-36">36</a> |
| 142 |
<a href="#section-7">7</a>. Operational Considerations .....................................<a href="#page-37">37</a> |
| 143 |
<a href="#section-7.1">7.1</a>. Configuration .............................................<a href="#page-37">37</a> |
| 144 |
<a href="#section-8">8</a>. IANA Considerations ............................................<a href="#page-37">37</a> |
| 145 |
<a href="#section-9">9</a>. Security Considerations ........................................<a href="#page-38">38</a> |
| 146 |
<a href="#section-10">10</a>. References ....................................................<a href="#page-39">39</a> |
| 147 |
<a href="#section-10.1">10.1</a>. Normative References .....................................<a href="#page-39">39</a> |
| 148 |
<a href="#section-10.2">10.2</a>. Informative References ...................................<a href="#page-39">39</a> |
| 149 |
Acknowledgements ..................................................<a href="#page-40">40</a> |
| 150 |
Contributors ......................................................<a href="#page-41">41</a> |
| 151 |
Authors' Addresses ................................................<a href="#page-43">43</a> |
| 152 |
|
| 153 |
|
| 154 |
|
| 155 |
|
| 156 |
|
| 157 |
|
| 158 |
|
| 159 |
|
| 160 |
|
| 161 |
|
| 162 |
|
| 163 |
|
| 164 |
<span class="grey">Wijnands, et al. Experimental [Page 3]</span></pre> |
| 165 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span> |
| 166 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 167 |
|
| 168 |
|
| 169 |
<span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span> |
| 170 |
|
| 171 |
This document specifies a new architecture for the forwarding of |
| 172 |
multicast data packets. The architecture provides optimal forwarding |
| 173 |
of multicast data packets through a "multicast domain". However, it |
| 174 |
does not require the use of a protocol for explicitly building |
| 175 |
multicast distribution trees, and it does not require intermediate |
| 176 |
nodes to maintain any per-flow state. This architecture is known as |
| 177 |
"Bit Index Explicit Replication" (BIER). |
| 178 |
|
| 179 |
A router that supports BIER is known as a "Bit-Forwarding Router" |
| 180 |
(BFR). The BIER control-plane protocols (see <a href="#section-4.2">Section 4.2</a>) run within |
| 181 |
a "BIER domain", allowing the BFRs within that domain to exchange the |
| 182 |
information needed for them to forward packets to each other |
| 183 |
using BIER. |
| 184 |
|
| 185 |
A multicast data packet enters a BIER domain at a "Bit-Forwarding |
| 186 |
Ingress Router" (BFIR), and leaves the BIER domain at one or more |
| 187 |
"Bit-Forwarding Egress Routers" (BFERs). A BFR that receives a |
| 188 |
multicast data packet from another BFR in the same BIER domain, and |
| 189 |
forwards the packet to another BFR in the same BIER domain, will be |
| 190 |
known as a "transit BFR" for that packet. A single BFR may be a BFIR |
| 191 |
for some multicast traffic while also being a BFER for some multicast |
| 192 |
traffic and a transit BFR for some multicast traffic. In fact, for a |
| 193 |
given packet, a BFR may be a BFIR and/or a transit BFR and/or (one |
| 194 |
of) the BFER(s) for that packet. |
| 195 |
|
| 196 |
A BIER domain may contain one or more sub-domains. Each BIER domain |
| 197 |
MUST contain at least one sub-domain, the "default sub-domain" (also |
| 198 |
denoted "sub-domain 0"). If a BIER domain contains more than one |
| 199 |
sub-domain, each BFR in the domain MUST be provisioned to know the |
| 200 |
set of sub-domains to which it belongs. Each sub-domain is |
| 201 |
identified by a sub-domain-id in the range [0,255]. |
| 202 |
|
| 203 |
For each sub-domain to which a given BFR belongs, if the BFR is |
| 204 |
capable of acting as a BFIR or a BFER, it MUST be provisioned with a |
| 205 |
"BFR-id" that is unique within the sub-domain. A BFR-id is a small |
| 206 |
unstructured positive integer. For instance, if a particular BIER |
| 207 |
sub-domain contains 1,374 BFRs, each one could be given a BFR-id in |
| 208 |
the range [1,1374]. |
| 209 |
|
| 210 |
If a given BFR belongs to more than one sub-domain, it may (though it |
| 211 |
need not) have a different BFR-id for each sub-domain. |
| 212 |
|
| 213 |
When a multicast packet arrives from outside the domain at a BFIR, |
| 214 |
the BFIR determines the set of BFERs to which the packet will be |
| 215 |
sent. The BFIR also determines the sub-domain in which the packet |
| 216 |
will be sent. Determining the sub-domain in which a given packet |
| 217 |
|
| 218 |
|
| 219 |
|
| 220 |
<span class="grey">Wijnands, et al. Experimental [Page 4]</span></pre> |
| 221 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span> |
| 222 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 223 |
|
| 224 |
|
| 225 |
will be sent is known as "assigning the packet to a sub-domain". |
| 226 |
Procedures for choosing the sub-domain to which a particular packet |
| 227 |
is assigned are outside the scope of this document. However, once a |
| 228 |
particular packet has been assigned to a particular sub-domain, it |
| 229 |
remains assigned to that sub-domain until it leaves the BIER domain. |
| 230 |
That is, the sub-domain to which a packet is assigned MUST NOT be |
| 231 |
changed while the packet is in flight through the BIER domain. |
| 232 |
|
| 233 |
Once the BFIR determines the sub-domain and the set of BFERs for a |
| 234 |
given packet, the BFIR encapsulates the packet in a "BIER header". |
| 235 |
The BIER header contains a bit string in which each bit represents a |
| 236 |
single BFR-id. To indicate that a particular BFER is to receive a |
| 237 |
given packet, the BFIR sets the bit corresponding to that BFER's |
| 238 |
BFR-id in the sub-domain to which the packet has been assigned. We |
| 239 |
will use the term "BitString" to refer to the bit string field in the |
| 240 |
BIER header. We will use the term "payload" to refer to the packet |
| 241 |
that has been encapsulated. Thus, a "BIER-encapsulated" packet |
| 242 |
consists of a "BIER header" followed by a "payload". |
| 243 |
|
| 244 |
The number of BFERs to which a given packet can be forwarded is |
| 245 |
limited only by the length of the BitString in the BIER header. |
| 246 |
Different deployments can use different BitString lengths. We will |
| 247 |
use the term "BitStringLength" to refer to the number of bits in the |
| 248 |
BitString. It is possible that some deployments will have more BFERs |
| 249 |
in a given sub-domain than there are bits in the BitString. To |
| 250 |
accommodate this case, the BIER encapsulation includes both the |
| 251 |
BitString and a "Set Identifier" (SI). It is the BitString and the |
| 252 |
SI together that determine the set of BFERs to which a given packet |
| 253 |
will be delivered: |
| 254 |
|
| 255 |
o By convention, the least significant (rightmost) bit in the |
| 256 |
BitString is "bit 1", and the most significant (leftmost) bit is |
| 257 |
"bit BitStringLength". |
| 258 |
|
| 259 |
o If a BIER-encapsulated packet has an SI of n and a BitString with |
| 260 |
bit k set, then the packet must be delivered to the BFER whose |
| 261 |
BFR-id (in the sub-domain to which the packet has been assigned) |
| 262 |
is n*BitStringLength+k. |
| 263 |
|
| 264 |
For example, suppose the BIER encapsulation uses a BitStringLength of |
| 265 |
256 bits. By convention, the least significant (rightmost) bit is |
| 266 |
bit 1, and the most significant (leftmost) bit is bit 256. Suppose |
| 267 |
that a given packet has been assigned to sub-domain 0 and needs to be |
| 268 |
delivered to three BFERs, where those BFERs have BFR-ids in |
| 269 |
sub-domain 0 of 13, 126, and 235, respectively. The BFIR would |
| 270 |
create a BIER encapsulation with the SI set to zero and with bits 13, |
| 271 |
126, and 235 of the BitString set. (All other bits of the BitString |
| 272 |
would be clear.) If the packet also needs to be sent to a BFER whose |
| 273 |
|
| 274 |
|
| 275 |
|
| 276 |
<span class="grey">Wijnands, et al. Experimental [Page 5]</span></pre> |
| 277 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span> |
| 278 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 279 |
|
| 280 |
|
| 281 |
BFR-id is 257, the BFIR would have to create a second copy of the |
| 282 |
packet, and the BIER encapsulation would specify an SI of 1, and a |
| 283 |
BitString with bit 1 set and all the other bits clear. |
| 284 |
|
| 285 |
It is generally advantageous to assign the BFR-ids of a given |
| 286 |
sub-domain so that as many BFERs as possible can be represented in a |
| 287 |
single bit string. |
| 288 |
|
| 289 |
Suppose a BFR (call it "BFR-A") receives a packet whose BIER |
| 290 |
encapsulation specifies an SI of 0 and a BitString with bits 13, 26, |
| 291 |
and 235 set. Suppose BFR-A has two BFR neighbors, BFR-B and BFR-C, |
| 292 |
such that the best path to BFERs 13 and 26 is via BFR-B, but the best |
| 293 |
path to BFER 235 is via BFR-C. BFR-A will then replicate the packet, |
| 294 |
sending one copy to BFR-B and one copy to BFR-C. However, BFR-A will |
| 295 |
clear bit 235 in the BitString of the packet copy it sends to BFR-B |
| 296 |
and will clear bits 13 and 26 in the BitString of the packet copy it |
| 297 |
sends to BFR-C. As a result, BFR-B will forward the packet only |
| 298 |
towards BFERs 13 and 26, and BFR-C will forward the packet only |
| 299 |
towards BFER 235. This ensures that each BFER receives only one copy |
| 300 |
of the packet. |
| 301 |
|
| 302 |
Detailed procedures for forwarding a BIER-encapsulated packet through |
| 303 |
a BIER domain can be found in <a href="#section-6">Section 6</a>. |
| 304 |
|
| 305 |
With this forwarding procedure, a multicast data packet can follow an |
| 306 |
optimal path from its BFIR to each of its BFERs. Further, since the |
| 307 |
set of BFERs for a given packet is explicitly encoded into the BIER |
| 308 |
header, the packet is not sent to any BFER that does not need to |
| 309 |
receive it. This allows for optimal forwarding of multicast traffic. |
| 310 |
This optimal forwarding is achieved without any need for transit BFRs |
| 311 |
to maintain per-flow state or to run a multicast tree-building |
| 312 |
protocol. |
| 313 |
|
| 314 |
The idea of encoding the set of egress nodes into the header of a |
| 315 |
multicast packet is not new. For example, [<a href="#ref-Boivie_Feldman">Boivie_Feldman</a>] proposes |
| 316 |
to encode the set of egress nodes as a set of IP addresses, and |
| 317 |
proposes mechanisms and procedures that are in some ways similar to |
| 318 |
those described in the current document. However, since BIER encodes |
| 319 |
each BFR-id as a single bit in a bit string, it can represent up to |
| 320 |
128 BFERs in the same number of bits that it would take to carry the |
| 321 |
IPv6 address of a single BFER. Thus, BIER scales to a much larger |
| 322 |
number of egress nodes per packet. |
| 323 |
|
| 324 |
BIER does not require that each transit BFR look up the best path to |
| 325 |
each BFER that is identified in the BIER header; the number of |
| 326 |
lookups required in the forwarding path for a single packet can be |
| 327 |
|
| 328 |
|
| 329 |
|
| 330 |
|
| 331 |
|
| 332 |
<span class="grey">Wijnands, et al. Experimental [Page 6]</span></pre> |
| 333 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span> |
| 334 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 335 |
|
| 336 |
|
| 337 |
limited to the number of neighboring BFRs; this can be much smaller |
| 338 |
than the number of BFERs. See <a href="#section-6">Section 6</a> (especially <a href="#section-6.5">Section 6.5</a>) for |
| 339 |
details. |
| 340 |
|
| 341 |
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", |
| 342 |
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and |
| 343 |
"OPTIONAL" in this document are to be interpreted as described in |
| 344 |
<a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a> [<a href="./rfc2119" title=""Key words for use in RFCs to Indicate Requirement Levels"">RFC2119</a>] [<a href="./rfc8174" title=""Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words"">RFC8174</a>] when, and only when, they appear in all |
| 345 |
capitals, as shown here. |
| 346 |
|
| 347 |
<span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. The BFR Identifier and BFR-Prefix</span> |
| 348 |
|
| 349 |
Each BFR MUST be assigned a single "BFR-prefix" for each sub-domain |
| 350 |
to which it belongs. A BFR's BFR-prefix MUST be an IP address |
| 351 |
(either IPv4 or IPv6) of the BFR. It is RECOMMENDED that the |
| 352 |
BFR-prefix be a loopback address of the BFR. |
| 353 |
|
| 354 |
If a BFR belongs to more than one sub-domain, it may (though it need |
| 355 |
not) have a different BFR-prefix in each sub-domain. |
| 356 |
|
| 357 |
All BFR-prefixes used within a given sub-domain MUST belong to the |
| 358 |
same address family (either IPv4 or IPv6). |
| 359 |
|
| 360 |
The BFR-prefix of a given BFR in a given sub-domain MUST be routable |
| 361 |
in that sub-domain. Whether a particular BFR-prefix is routable in a |
| 362 |
given sub-domain depends on the "routing underlay" associated with |
| 363 |
that sub-domain. The notion of "routing underlay" is described in |
| 364 |
<a href="#section-4.1">Section 4.1</a>. |
| 365 |
|
| 366 |
A "BFR Identifier" (BFR-id) is a number in the range [1,65535]. |
| 367 |
Within a given sub-domain, every BFR that may need to function as a |
| 368 |
BFIR or BFER MUST have a single BFR-id, which identifies it uniquely |
| 369 |
within that sub-domain. A BFR that does not need to function as a |
| 370 |
BFIR or BFER in a given sub-domain does not need to have a BFR-id in |
| 371 |
that sub-domain. |
| 372 |
|
| 373 |
The value 0 is not a legal BFR-id. |
| 374 |
|
| 375 |
The procedure for assigning a particular BFR-id to a particular BFR |
| 376 |
is outside the scope of this document. However, it is RECOMMENDED |
| 377 |
that the BFR-ids for each sub-domain be assigned "densely" from the |
| 378 |
numbering space, as this will result in a more efficient encoding |
| 379 |
(see <a href="#section-3">Section 3</a>). That is, if there are 256 or fewer BFERs, it is |
| 380 |
RECOMMENDED to assign all the BFR-ids from the range [1,256]. If |
| 381 |
there are more than 256 BFERs but less than 512, it is RECOMMENDED to |
| 382 |
assign all the BFR-ids from the range [1,512], with as few "holes" as |
| 383 |
|
| 384 |
|
| 385 |
|
| 386 |
|
| 387 |
|
| 388 |
<span class="grey">Wijnands, et al. Experimental [Page 7]</span></pre> |
| 389 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-8" ></span> |
| 390 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 391 |
|
| 392 |
|
| 393 |
possible in the earlier range. However, in some deployments, it may |
| 394 |
be advantageous to depart from this recommendation; this is discussed |
| 395 |
further in <a href="#section-3">Section 3</a>. |
| 396 |
|
| 397 |
In some deployments, it may not be possible to support (in a given |
| 398 |
sub-domain) the full range of 65,535 BFR-ids. For example, if the |
| 399 |
BFRs in a given sub-domain only support 16 SIs and if they only |
| 400 |
support BitStringLengths of 256 or less, then only 16*256=4,096 |
| 401 |
BFR-ids can be supported in that sub-domain. |
| 402 |
|
| 403 |
<span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Encoding BFR Identifiers in BitStrings</span> |
| 404 |
|
| 405 |
To encode a BFR-id in a BIER data packet, one must convert the BFR-id |
| 406 |
to an SI and a BitString. This conversion depends upon the parameter |
| 407 |
we are calling "BitStringLength". The conversion is done as follows. |
| 408 |
If the BFR-id is N, then |
| 409 |
|
| 410 |
o SI is the integer part of the quotient (N-1)/BitStringLength. |
| 411 |
|
| 412 |
o The BitString has one bit position set. If the low-order bit is |
| 413 |
bit 1 and the high-order bit is bit BitStringLength, the bit |
| 414 |
position that represents BFR-id N is |
| 415 |
((N-1) modulo BitStringLength)+1. |
| 416 |
|
| 417 |
If several different BFR-ids all resolve to the same SI, then all of |
| 418 |
those BFR-ids can be represented in a single BitString. The |
| 419 |
BitStrings for all of those BFR-ids are combined using a bitwise |
| 420 |
logical OR operation. |
| 421 |
|
| 422 |
Within a given BIER domain (or even within a given BIER sub-domain), |
| 423 |
different values of BitStringLength may be used. Each BFR MUST be |
| 424 |
provisioned to know the following: |
| 425 |
|
| 426 |
o The BitStringLength ("Imposition BitStringLength") and sub-domain |
| 427 |
("Imposition sub-domain") to use when it imposes (as a BFIR) a |
| 428 |
BIER encapsulation on a particular set of packets, and |
| 429 |
|
| 430 |
o The BitStringLengths ("Disposition BitStringLengths") that it will |
| 431 |
process when (as a BFR or BFER) it receives packets from a |
| 432 |
particular sub-domain. |
| 433 |
|
| 434 |
It is not required that a BFIR use the same Imposition |
| 435 |
BitStringLength or the same Imposition sub-domain for all packets on |
| 436 |
which it imposes the BIER encapsulation. However, if a particular |
| 437 |
BFIR is provisioned to use a particular Imposition BitStringLength |
| 438 |
and a particular Imposition sub-domain when imposing the |
| 439 |
encapsulation on a given set of packets, all other BFRs with BFR-ids |
| 440 |
in that sub-domain SHOULD be provisioned to process received BIER |
| 441 |
|
| 442 |
|
| 443 |
|
| 444 |
<span class="grey">Wijnands, et al. Experimental [Page 8]</span></pre> |
| 445 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-9" ></span> |
| 446 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 447 |
|
| 448 |
|
| 449 |
packets with that BitStringLength (i.e., all other BFRs with BFR-ids |
| 450 |
in that sub-domain SHOULD be provisioned with that BitStringLength as |
| 451 |
a Disposition BitStringLength for that sub-domain). Exceptions to |
| 452 |
this rule MAY be made under certain conditions; this is discussed in |
| 453 |
<a href="#section-6.10">Section 6.10</a>. |
| 454 |
|
| 455 |
When a BIER encapsulation is specified, the specification MUST define |
| 456 |
a default BitStringLength for the encapsulation. Every BFIR |
| 457 |
supporting that encapsulation MUST be capable of being provisioned |
| 458 |
with that default BitStringLength as its Imposition BitStringLength. |
| 459 |
Every BFR and BFER supporting that encapsulation MUST be capable of |
| 460 |
being provisioned with that default BitStringLength as a Disposition |
| 461 |
BitStringLength. |
| 462 |
|
| 463 |
The specification of a BIER encapsulation MAY also allow the use of |
| 464 |
other BitStringLengths. |
| 465 |
|
| 466 |
If a BFR is capable of being provisioned with a given value of |
| 467 |
BitStringLength as an Imposition BitStringLength, it MUST also be |
| 468 |
capable of being provisioned with that same value as one of its |
| 469 |
Disposition BitStringLengths. It SHOULD be capable of being |
| 470 |
provisioned with each legal smaller value of BitStringLength as (a) |
| 471 |
its Imposition BitStringLength, and (b) one of its Disposition |
| 472 |
BitStringLengths. |
| 473 |
|
| 474 |
In order to support transition from one BitStringLength to another, |
| 475 |
every BFR MUST be capable of being provisioned to simultaneously use |
| 476 |
two different Disposition BitStringLengths. |
| 477 |
|
| 478 |
A BFR MUST support SI values in the range [0,15] and MAY support SI |
| 479 |
values in the range [0,255]. ("Supporting the values in a given |
| 480 |
range" means, in this context, that any value in the given range is |
| 481 |
legal and will be properly interpreted.) Note that for a given |
| 482 |
BitStringLength, the total number of BFR-ids that can be represented |
| 483 |
is the product of the BitStringLength and the number of supported |
| 484 |
SIs. For example, if a deployment uses (in a given sub-domain) a |
| 485 |
BitStringLength of 64 and supports 256 SIs, that deployment can only |
| 486 |
support 16384 BFR-ids in that sub-domain. Even a deployment that |
| 487 |
supports 256 SIs will not be able to support 65,535 BFR-ids unless it |
| 488 |
uses a BitStringLength of at least 256. |
| 489 |
|
| 490 |
When a BFIR determines that a multicast data packet, assigned to a |
| 491 |
given sub-domain, needs to be forwarded to a particular set of |
| 492 |
destination BFERs, the BFIR partitions that set of BFERs into |
| 493 |
subsets, where each subset contains the target BFERs whose BFR-ids in |
| 494 |
the given sub-domain all resolve to the same SI. Call these the |
| 495 |
"SI-subsets" for the packet. Each SI-subset can be represented by a |
| 496 |
single BitString. The BFIR creates a copy of the packet for each |
| 497 |
|
| 498 |
|
| 499 |
|
| 500 |
<span class="grey">Wijnands, et al. Experimental [Page 9]</span></pre> |
| 501 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-10" ></span> |
| 502 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 503 |
|
| 504 |
|
| 505 |
SI-subset. The BIER encapsulation is then applied to each packet. |
| 506 |
The encapsulation specifies a single SI for each packet and contains |
| 507 |
the BitString that represents all the BFR-ids in the corresponding |
| 508 |
SI-subset. Of course, in order to properly interpret the BitString, |
| 509 |
it must be possible to infer the sub-domain-id from the encapsulation |
| 510 |
as well. |
| 511 |
|
| 512 |
Suppose, for example, that a BFIR determines that a given packet |
| 513 |
needs to be forwarded to three BFERs, whose BFR-ids (in the |
| 514 |
appropriate sub-domain) are 27, 235, and 497. The BFIR will have to |
| 515 |
forward two copies of the packet. One copy, associated with SI=0, |
| 516 |
will have a BitString with bits 27 and 235 set. The other copy, |
| 517 |
associated with SI=1, will have a BitString with bit 241 set. |
| 518 |
|
| 519 |
In order to minimize the number of copies that must be made of a |
| 520 |
given multicast packet, it is RECOMMENDED that the BFR-ids used in a |
| 521 |
given sub-domain be assigned "densely" (see <a href="#section-2">Section 2</a>) from the |
| 522 |
numbering space. This will minimize the number of SIs that have to |
| 523 |
be used in that sub-domain. However, depending upon the details of a |
| 524 |
particular deployment, other assignment methods may be more |
| 525 |
advantageous. Suppose, for example, that in a certain deployment, |
| 526 |
every multicast flow is intended either for the "east coast" or for |
| 527 |
the "west coast", but not for both coasts. In such a deployment, it |
| 528 |
would be advantageous to assign BFR-ids so that all the "west coast" |
| 529 |
BFR-ids fall into the same SI-subset and so that all the "east coast" |
| 530 |
BFR-ids fall into the same SI-subset. |
| 531 |
|
| 532 |
When a BFR receives a BIER data packet, it will infer the SI from the |
| 533 |
encapsulation. The set of BFERs to which the packet needs to be |
| 534 |
forwarded can then be inferred from the SI and the BitString. |
| 535 |
|
| 536 |
In some of the examples given later in this document, we will use a |
| 537 |
BitStringLength of 4 and will represent a BFR-id in the form |
| 538 |
"SI:xyzw", where SI is the Set Identifier of the BFR-id (assuming a |
| 539 |
BitStringLength of 4) and xyzw is a string of 4 bits. A |
| 540 |
BitStringLength of 4 is used only in the examples; we would not |
| 541 |
expect actual deployments to have such a small BitStringLength. |
| 542 |
|
| 543 |
It is possible that several different forms of BIER encapsulation |
| 544 |
will be developed. If so, the particular encapsulation that is used |
| 545 |
in a given deployment will depend on the type of network |
| 546 |
infrastructure that is used to realize the BIER domain. Details of |
| 547 |
the BIER encapsulation(s) will be given in companion documents. An |
| 548 |
encapsulation for use in MPLS networks is described in |
| 549 |
[<a href="#ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>]; that document also describes a very similar |
| 550 |
encapsulation that can be used in non-MPLS networks. |
| 551 |
|
| 552 |
|
| 553 |
|
| 554 |
|
| 555 |
|
| 556 |
<span class="grey">Wijnands, et al. Experimental [Page 10]</span></pre> |
| 557 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-11" ></span> |
| 558 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 559 |
|
| 560 |
|
| 561 |
<span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. Layering</span> |
| 562 |
|
| 563 |
It is helpful to think of the BIER architecture as consisting of |
| 564 |
three layers: the "routing underlay", the "BIER layer", and the |
| 565 |
"multicast flow overlay". |
| 566 |
|
| 567 |
<span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. The Routing Underlay</span> |
| 568 |
|
| 569 |
The "routing underlay" establishes "adjacencies" between pairs of |
| 570 |
BFRs and determines one or more "best paths" from a given BFR to a |
| 571 |
given set of BFRs. Each such path is a sequence of BFRs |
| 572 |
<BFR(k), BFR(k+1), ..., BFR(k+n)> such that BFR(k+j) is "adjacent" to |
| 573 |
BFR(k+j+1) (for 0<=j<n). |
| 574 |
|
| 575 |
At a given BFR, say BFR-A, for every IP address that is the address |
| 576 |
of a BFR in the BIER domain, the routing underlay will map that IP |
| 577 |
address into a set of one or more "equal-cost" adjacencies. If a |
| 578 |
BIER data packet has to be forwarded by BFR-A to a given BFER, say |
| 579 |
BFER-B, the packet will follow the path from BFR-A to BFER-B that is |
| 580 |
determined by the routing underlay. |
| 581 |
|
| 582 |
It is expected that in a typical deployment, the routing underlay |
| 583 |
will be the default topology that the Interior Gateway Protocol |
| 584 |
(IGP), e.g., OSPF, uses for unicast routing. In that case, the |
| 585 |
underlay adjacencies are just the OSPF adjacencies. A BIER data |
| 586 |
packet traveling from BFR-A to BFER-B will follow the path that OSPF |
| 587 |
has selected for unicast traffic from BFR-A to BFER-B. |
| 588 |
|
| 589 |
If one wants to have multicast traffic from BFR-A to BFER-B travel a |
| 590 |
path that is different from the path used by the unicast traffic from |
| 591 |
BFR-A to BFER-B, one can use a different underlay. For example, if |
| 592 |
multi-topology OSPF is being used, one OSPF topology could be used |
| 593 |
for unicast traffic and the other for multicast traffic. (Each |
| 594 |
topology would be considered to be a different underlay.) |
| 595 |
Alternatively, one could deploy a routing underlay that creates a |
| 596 |
multicast-specific tree of some sort. BIER could then be used to |
| 597 |
forward multicast data packets along the multicast-specific tree, |
| 598 |
while unicast packets follow the "ordinary" OSPF best path. (In a |
| 599 |
case like this, many multicast flows could be traveling along a |
| 600 |
single tree, and the BitString carried by a particular packet would |
| 601 |
identify those nodes of the tree that need to receive that packet.) |
| 602 |
It is even possible to have multiple routing underlays used by BIER, |
| 603 |
as long as one can infer from a data packet's BIER encapsulation |
| 604 |
which underlay is being used for that packet. |
| 605 |
|
| 606 |
|
| 607 |
|
| 608 |
|
| 609 |
|
| 610 |
|
| 611 |
|
| 612 |
<span class="grey">Wijnands, et al. Experimental [Page 11]</span></pre> |
| 613 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-12" ></span> |
| 614 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 615 |
|
| 616 |
|
| 617 |
If multiple routing underlays are used in a single BIER domain, each |
| 618 |
BIER sub-domain MUST be associated with a single routing underlay |
| 619 |
(though multiple sub-domains may be associated with the same routing |
| 620 |
underlay). A BFR that belongs to multiple sub-domains MUST be |
| 621 |
provisioned to know which routing underlay is used by each |
| 622 |
sub-domain. By default (i.e., in the absence of any provisioning to |
| 623 |
the contrary), each sub-domain uses the default topology of the |
| 624 |
unicast IGP as the routing underlay. |
| 625 |
|
| 626 |
In scenarios where External BGP (EBGP) is used as the IGP, the |
| 627 |
underlay adjacencies, by default, are the BGP adjacencies. |
| 628 |
|
| 629 |
Specification of the protocols and procedures of the routing underlay |
| 630 |
is outside the scope of this document. |
| 631 |
|
| 632 |
<span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a>. The BIER Layer</span> |
| 633 |
|
| 634 |
The BIER layer consists of the protocols and procedures that are used |
| 635 |
in order to transmit a multicast data packet across a BIER domain, |
| 636 |
from its BFIR to its BFERs. This includes the following components: |
| 637 |
|
| 638 |
o Protocols and procedures that a given BFR uses to advertise, to |
| 639 |
all other BFRs in the same BIER domain: |
| 640 |
|
| 641 |
* its BFR-prefix; |
| 642 |
|
| 643 |
* its BFR-id in each sub-domain for which it has been provisioned |
| 644 |
with a BFR-id; |
| 645 |
|
| 646 |
* the set of Disposition BitStringLengths it has been provisioned |
| 647 |
to use for each sub-domain; |
| 648 |
|
| 649 |
* optionally, information about the routing underlay associated |
| 650 |
with each sub-domain. |
| 651 |
|
| 652 |
o The procedures used by a BFIR to impose a BIER header on a |
| 653 |
multicast data packet. |
| 654 |
|
| 655 |
o The procedures for forwarding BIER-encapsulated packets and for |
| 656 |
modifying the BIER header during transit. |
| 657 |
|
| 658 |
o The procedures used by a BFER to decapsulate a BIER packet and |
| 659 |
properly dispatch it. |
| 660 |
|
| 661 |
|
| 662 |
|
| 663 |
|
| 664 |
|
| 665 |
|
| 666 |
|
| 667 |
|
| 668 |
<span class="grey">Wijnands, et al. Experimental [Page 12]</span></pre> |
| 669 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-13" ></span> |
| 670 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 671 |
|
| 672 |
|
| 673 |
<span class="h3"><a class="selflink" id="section-4.3" href="#section-4.3">4.3</a>. The Multicast Flow Overlay</span> |
| 674 |
|
| 675 |
The "multicast flow overlay" consists of the set of protocols and |
| 676 |
procedures that enable the following set of functions. |
| 677 |
|
| 678 |
o When a BFIR receives a multicast data packet from outside the BIER |
| 679 |
domain, the BFIR must determine the set of BFERs for that packet. |
| 680 |
This information is provided by the multicast flow overlay. |
| 681 |
|
| 682 |
o When a BFER receives a BIER-encapsulated packet from inside the |
| 683 |
BIER domain, the BFER must determine how to further forward the |
| 684 |
packet. This information is provided by the multicast flow |
| 685 |
overlay. |
| 686 |
|
| 687 |
For example, suppose the BFIR and BFERs are Provider Edge (PE) |
| 688 |
routers providing Multicast Virtual Private Network (MVPN) service. |
| 689 |
The multicast flow overlay consists of the protocols and procedures |
| 690 |
described in [<a href="./rfc6513" title=""Multicast in MPLS/BGP IP VPNs"">RFC6513</a>] and [<a href="./rfc6514" title=""BGP Encodings and Procedures for Multicast in MPLS/BGP IP VPNs"">RFC6514</a>]. The MVPN signaling described |
| 691 |
in those RFCs enables an ingress PE to determine the set of egress |
| 692 |
PEs for a given multicast flow (or set of flows); it also enables an |
| 693 |
egress PE to determine the "Virtual Routing and Forwarding Tables" |
| 694 |
(VRFs) to which multicast packets from the backbone network should be |
| 695 |
sent. MVPN signaling also has several components that depend on the |
| 696 |
type of "tunneling technology" used to carry multicast data through |
| 697 |
the network. Since BIER is, in effect, a new type of "tunneling |
| 698 |
technology", some extensions to the MVPN signaling are needed in |
| 699 |
order to properly interface the multicast flow overlay with the BIER |
| 700 |
layer. These are specified in [<a href="#ref-BIER_MVPN">BIER_MVPN</a>]. |
| 701 |
|
| 702 |
MVPN is just one example of a multicast flow overlay. Protocols and |
| 703 |
procedures for other overlays will be provided in companion |
| 704 |
documents. It is also possible to implement the multicast flow |
| 705 |
overlay by means of a "Software-Defined Network" (SDN) controller. |
| 706 |
Specification of the protocols and procedures of the multicast flow |
| 707 |
overlay is outside the scope of this document. |
| 708 |
|
| 709 |
<span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. Advertising BFR-ids and BFR-Prefixes</span> |
| 710 |
|
| 711 |
As stated in <a href="#section-2">Section 2</a>, each BFER is assigned (by provisioning) a |
| 712 |
BFR-id (for a given BIER sub-domain). Each BFER must advertise these |
| 713 |
assignments to all the other BFRs in the domain. Similarly, each BFR |
| 714 |
is assigned (by provisioning) a BFR-prefix (for a given BIER domain) |
| 715 |
and must advertise this assignment to all the other BFRs in the |
| 716 |
domain. Finally, each BFR has been provisioned to use a certain set |
| 717 |
of Disposition BitStringLengths for each sub-domain and must |
| 718 |
advertise these to all other BFRs in the domain. |
| 719 |
|
| 720 |
|
| 721 |
|
| 722 |
|
| 723 |
|
| 724 |
<span class="grey">Wijnands, et al. Experimental [Page 13]</span></pre> |
| 725 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-14" ></span> |
| 726 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 727 |
|
| 728 |
|
| 729 |
If the BIER domain is also a link-state routing IGP domain (i.e., an |
| 730 |
OSPF or IS-IS domain), the advertisement of the BFR-prefix, |
| 731 |
<sub-domain-id, BFR-id>, and BitStringLength can be done using the |
| 732 |
advertisement capabilities of the IGP. For example, if a BIER domain |
| 733 |
is also an OSPF domain, these advertisements can be done using the |
| 734 |
OSPF "Opaque Link State Advertisement" (Opaque LSA) mechanism. |
| 735 |
Details of the necessary extensions to OSPF and IS-IS will be |
| 736 |
provided in companion documents. (See [<a href="#ref-OSPF_BIER_EXTENSIONS">OSPF_BIER_EXTENSIONS</a>] and |
| 737 |
[<a href="#ref-ISIS_BIER_EXTENSIONS">ISIS_BIER_EXTENSIONS</a>].) |
| 738 |
|
| 739 |
If, in a particular deployment, the BIER domain is not an OSPF or |
| 740 |
IS-IS domain, procedures suitable to the deployment must be used to |
| 741 |
advertise this information. Details of the necessary procedures will |
| 742 |
be provided in companion documents. For example, if BGP is the only |
| 743 |
routing algorithm used in the BIER domain, the procedures of |
| 744 |
[<a href="#ref-BGP_BIER_EXTENSIONS">BGP_BIER_EXTENSIONS</a>] may be used. |
| 745 |
|
| 746 |
These advertisements enable each BFR to associate a given |
| 747 |
<sub-domain-id, BFR-id> with a given BFR-prefix. As will be seen in |
| 748 |
subsequent sections of this document, knowledge of this association |
| 749 |
is an important part of the forwarding process. |
| 750 |
|
| 751 |
Since each BFR needs to have a unique (in each sub-domain) BFR-id, |
| 752 |
two different BFRs will not advertise ownership of the same |
| 753 |
<sub-domain-id, BFR-id> unless there has been a provisioning error. |
| 754 |
|
| 755 |
o If BFR-A determines that BFR-B and BFR-C have both advertised the |
| 756 |
same BFR-id for the same sub-domain, BFR-A MUST log an error. |
| 757 |
Suppose that the duplicate BFR-id is "N". When BFR-A is |
| 758 |
functioning as a BFIR, it MUST NOT encode the BFR-id value N in |
| 759 |
the BIER encapsulation of any packet that has been assigned to the |
| 760 |
given sub-domain, even if it has determined that the packet needs |
| 761 |
to be received by BFR-B and/or BFR-C. |
| 762 |
|
| 763 |
This will mean that BFR-B and BFR-C cannot receive multicast |
| 764 |
traffic at all in the given sub-domain until the provisioning |
| 765 |
error is fixed. However, that is preferable to having them |
| 766 |
receive each other's traffic. |
| 767 |
|
| 768 |
o Suppose that BFR-A has been provisioned with BFR-id N for a |
| 769 |
particular sub-domain but that it has not yet advertised its |
| 770 |
ownership of BFR-id N for that sub-domain. Suppose also that it |
| 771 |
has received an advertisement from a different BFR (say BFR-B) |
| 772 |
that is advertising ownership of BFR-id N for the same sub-domain. |
| 773 |
In such a case, BFR-A SHOULD log an error and MUST NOT advertise |
| 774 |
its own ownership of BFR-id N for that sub-domain as long as the |
| 775 |
advertisement from BFR-B is extant. |
| 776 |
|
| 777 |
|
| 778 |
|
| 779 |
|
| 780 |
<span class="grey">Wijnands, et al. Experimental [Page 14]</span></pre> |
| 781 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-15" ></span> |
| 782 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 783 |
|
| 784 |
|
| 785 |
This procedure may prevent the accidental misconfiguration of a |
| 786 |
new BFR from impacting an existing BFR. |
| 787 |
|
| 788 |
If a BFR advertises that it has a BFR-id of 0 in a particular |
| 789 |
sub-domain, other BFRs receiving the advertisement MUST interpret |
| 790 |
that advertisement as meaning that the advertising BFR does not have |
| 791 |
a BFR-id in that sub-domain. |
| 792 |
|
| 793 |
<span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. BIER Intra-Domain Forwarding Procedures</span> |
| 794 |
|
| 795 |
This section specifies the rules for forwarding a BIER-encapsulated |
| 796 |
data packet within a BIER domain. These rules are not intended to |
| 797 |
specify an implementation strategy; to conform to this specification, |
| 798 |
an implementation need only produce the same results that these rules |
| 799 |
produce. |
| 800 |
|
| 801 |
<span class="h3"><a class="selflink" id="section-6.1" href="#section-6.1">6.1</a>. Overview</span> |
| 802 |
|
| 803 |
This section provides a brief overview of the BIER forwarding |
| 804 |
procedures. Subsequent subsections specify the procedures in more |
| 805 |
detail. |
| 806 |
|
| 807 |
To forward a BIER-encapsulated packet: |
| 808 |
|
| 809 |
1. Determine the packet's sub-domain. |
| 810 |
|
| 811 |
2. Determine the packet's BitStringLength and BitString. |
| 812 |
|
| 813 |
3. Determine the packet's SI. |
| 814 |
|
| 815 |
4. From the sub-domain, the SI, and the BitString, determine the set |
| 816 |
of destination BFERs for the packet. |
| 817 |
|
| 818 |
5. Using information provided by the routing underlay associated |
| 819 |
with the packet's sub-domain, determine the next-hop adjacency |
| 820 |
for each of the destination BFERs. |
| 821 |
|
| 822 |
6. It is possible that the packet's BitString will have one or more |
| 823 |
bits that correspond to BFR-ids that are not in use. It is also |
| 824 |
possible that the packet's BitString will have one or more bits |
| 825 |
that correspond to BFERs that are unreachable, i.e., that have no |
| 826 |
next-hop adjacency. In the following, we will consider the |
| 827 |
"next-hop adjacency" for all such bit positions to be the "null" |
| 828 |
next hop. |
| 829 |
|
| 830 |
7. Partition the set of destination BFERs such that all the BFERs in |
| 831 |
a single partition have the same next hop. We will say that each |
| 832 |
partition is associated with a next hop. |
| 833 |
|
| 834 |
|
| 835 |
|
| 836 |
<span class="grey">Wijnands, et al. Experimental [Page 15]</span></pre> |
| 837 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-16" ></span> |
| 838 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 839 |
|
| 840 |
|
| 841 |
8. For each partition: |
| 842 |
|
| 843 |
a. Make a copy of the packet. |
| 844 |
|
| 845 |
b. Clear any bit in the packet's BitString that identifies a |
| 846 |
BFER that is not in the partition. |
| 847 |
|
| 848 |
c. Transmit the packet to the associated next hop. (If the |
| 849 |
next hop is the null next hop, the packet is discarded.) |
| 850 |
|
| 851 |
If a BFR receives a BIER-encapsulated packet whose <sub-domain, SI, |
| 852 |
BitString> triple identifies that BFR itself, then the BFR is also a |
| 853 |
BFER for that packet. As a BFER, it must pass the payload to the |
| 854 |
multicast flow overlay. If the BitString has bits set for other |
| 855 |
BFRs, the packet also needs to be forwarded further within the BIER |
| 856 |
domain. If the BF(E)R also forwards one or more copies of the packet |
| 857 |
within the BIER domain, the bit representing the BFR's own BFR-id |
| 858 |
MUST be clear in all the copies. |
| 859 |
|
| 860 |
When BIER on a BFER is to pass a packet to the multicast flow |
| 861 |
overlay, it of course decapsulates the packet by removing the BIER |
| 862 |
header. However, it may be necessary to provide the multicast flow |
| 863 |
overlay with contextual information obtained from the BIER |
| 864 |
encapsulation. The information that needs to pass between the BIER |
| 865 |
layer and the multicast flow overlay is specific to the multicast |
| 866 |
flow overlay. Specification of the interaction between the BIER |
| 867 |
layer and the multicast flow overlay is outside the scope of this |
| 868 |
specification. |
| 869 |
|
| 870 |
If the BIER encapsulation contains a "Time to Live" (TTL) value, this |
| 871 |
value is not, by default, inherited by the payload. If a particular |
| 872 |
multicast flow overlay needs to know the TTL value, this needs to be |
| 873 |
specified in whatever specification defines the interaction between |
| 874 |
BIER and that multicast flow overlay. |
| 875 |
|
| 876 |
If the BIER encapsulation contains a Traffic Class field, a |
| 877 |
Type of Service field, a Differentiated Services field, or any field |
| 878 |
of that sort, the value of that field is not, by default, passed to |
| 879 |
the multicast flow overlay. If a particular multicast flow overlay |
| 880 |
needs to know the values of such fields, this fact needs to be |
| 881 |
specified in whatever specification defines the interaction between |
| 882 |
BIER and that multicast flow overlay. |
| 883 |
|
| 884 |
When BIER on a BFER passes a packet to the multicast flow overlay, |
| 885 |
the overlay will determine how to further dispatch the packet. If |
| 886 |
the packet needs to be forwarded into another BIER domain, then the |
| 887 |
BFR will act as a BFER in one BIER domain and as a BFIR in another. |
| 888 |
|
| 889 |
|
| 890 |
|
| 891 |
|
| 892 |
<span class="grey">Wijnands, et al. Experimental [Page 16]</span></pre> |
| 893 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-17" ></span> |
| 894 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 895 |
|
| 896 |
|
| 897 |
A BIER-encapsulated packet cannot pass directly from one BIER domain |
| 898 |
to another; at the boundary between BIER domains, the packet must be |
| 899 |
decapsulated and passed to the multicast flow overlay. |
| 900 |
|
| 901 |
Note that when a BFR transmits multiple copies of a packet within a |
| 902 |
BIER domain, only one copy will be destined to any given BFER. |
| 903 |
Therefore, it is not possible for any BIER-encapsulated packet to be |
| 904 |
delivered more than once to any BFER. |
| 905 |
|
| 906 |
<span class="h3"><a class="selflink" id="section-6.2" href="#section-6.2">6.2</a>. BFR Neighbors</span> |
| 907 |
|
| 908 |
The "BFR Neighbors" (BFR-NBRs) of a given BFR, say BFR-A, are those |
| 909 |
BFRs that, according to the routing underlay, are adjacencies of |
| 910 |
BFR-A. Each BFR-NBR will have a BFR-prefix. |
| 911 |
|
| 912 |
Suppose a BIER-encapsulated packet arrives at BFR-A. From the |
| 913 |
packet's encapsulation, BFR-A learns (a) the sub-domain of the packet |
| 914 |
and (b) the BFR-ids (in that sub-domain) of the BFERs to which the |
| 915 |
packet is destined. Then, using the information advertised per |
| 916 |
<a href="#section-5">Section 5</a>, BFR-A can find the BFR-prefix of each destination BFER. |
| 917 |
Given the BFR-prefix of a particular destination BFER, say BFER-D, |
| 918 |
BFR-A learns from the routing underlay (associated with the packet's |
| 919 |
sub-domain) an IP address of the BFR that is the next hop on the path |
| 920 |
from BFR-A to BFER-D. Let's call this next hop "BFR-B". BFR-A must |
| 921 |
then determine the BFR-prefix of BFR-B. (This determination can be |
| 922 |
made from the information advertised per <a href="#section-5">Section 5</a>.) This BFR-prefix |
| 923 |
is the BFR-NBR of BFR-A on the path from BFR-A to BFER-D. |
| 924 |
|
| 925 |
Note that if the routing underlay provides multiple equal-cost paths |
| 926 |
from BFR-A to BFER-D, BFR-A may have multiple BFR-NBRs for BFER-D. |
| 927 |
|
| 928 |
Under certain circumstances, a BFR may have adjacencies (in a |
| 929 |
particular routing underlay) that are not BFRs. Please see |
| 930 |
<a href="#section-6.9">Section 6.9</a> for a discussion of how to handle those circumstances. |
| 931 |
|
| 932 |
|
| 933 |
|
| 934 |
|
| 935 |
|
| 936 |
|
| 937 |
|
| 938 |
|
| 939 |
|
| 940 |
|
| 941 |
|
| 942 |
|
| 943 |
|
| 944 |
|
| 945 |
|
| 946 |
|
| 947 |
|
| 948 |
<span class="grey">Wijnands, et al. Experimental [Page 17]</span></pre> |
| 949 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-18" ></span> |
| 950 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 951 |
|
| 952 |
|
| 953 |
<span class="h3"><a class="selflink" id="section-6.3" href="#section-6.3">6.3</a>. The Bit Index Routing Table</span> |
| 954 |
|
| 955 |
The "Bit Index Routing Table" (BIRT) is a table that maps from the |
| 956 |
BFR-id (in a particular sub-domain) of a BFER to the BFR-prefix of |
| 957 |
that BFER, and to the BFR-NBR on the path to that BFER. As an |
| 958 |
example, consider the topology shown in Figure 1. In this diagram, |
| 959 |
we represent the BFR-id of each BFR in the SI:xyzw form discussed in |
| 960 |
<a href="#section-3">Section 3</a>. |
| 961 |
|
| 962 |
( A ) ------------ ( B ) ------------ ( C ) ------------ ( D ) |
| 963 |
4 (0:1000) \ \ 1 (0:0001) |
| 964 |
\ \ |
| 965 |
( E ) ( F ) |
| 966 |
3 (0:0100) 2 (0:0010) |
| 967 |
|
| 968 |
Figure 1: BIER Topology 1 |
| 969 |
|
| 970 |
This topology will result in the BIRT of Figure 2 at BFR-B. The |
| 971 |
first column shows the BFR-id as a number and also (in parentheses) |
| 972 |
in the SI:BitString format that corresponds to a BitStringLength |
| 973 |
of 4. (The actual minimum BitStringLength is 64, but we use 4 in the |
| 974 |
examples.) |
| 975 |
|
| 976 |
Note that a BIRT is specific to a particular BIER sub-domain. |
| 977 |
|
| 978 |
-------------------------------------------- |
| 979 |
| BFR-id | BFR-Prefix | BFR-NBR | |
| 980 |
| (SI:BitString) | of Dest BFER | | |
| 981 |
============================================ |
| 982 |
| 4 (0:1000) | A | A | |
| 983 |
-------------------------------------------- |
| 984 |
| 1 (0:0001) | D | C | |
| 985 |
-------------------------------------------- |
| 986 |
| 3 (0:0100) | E | E | |
| 987 |
-------------------------------------------- |
| 988 |
| 2 (0:0010) | F | C | |
| 989 |
-------------------------------------------- |
| 990 |
|
| 991 |
Figure 2: Bit Index Routing Table at BFR-B |
| 992 |
|
| 993 |
|
| 994 |
|
| 995 |
|
| 996 |
|
| 997 |
|
| 998 |
|
| 999 |
|
| 1000 |
|
| 1001 |
|
| 1002 |
|
| 1003 |
|
| 1004 |
<span class="grey">Wijnands, et al. Experimental [Page 18]</span></pre> |
| 1005 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-19" ></span> |
| 1006 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1007 |
|
| 1008 |
|
| 1009 |
<span class="h3"><a class="selflink" id="section-6.4" href="#section-6.4">6.4</a>. The Bit Index Forwarding Table</span> |
| 1010 |
|
| 1011 |
The "Bit Index Forwarding Table" (BIFT) is derived from the BIRT as |
| 1012 |
follows. (Note that a BIFT is specific to a particular sub-domain.) |
| 1013 |
|
| 1014 |
Suppose that several rows in the BIRT have the same SI and the same |
| 1015 |
BFR-NBR. By taking the logical OR of the BitStrings of those rows, |
| 1016 |
we obtain a bit mask that corresponds to that combination of SI and |
| 1017 |
BFR-NBR. We will refer to this bit mask as the "Forwarding Bit Mask" |
| 1018 |
(F-BM) for that <SI, BFR-NBR> combination. |
| 1019 |
|
| 1020 |
For example, in Figure 2, we see that two of the rows have the same |
| 1021 |
SI (0) and same BFR-NBR (C). The bit mask that corresponds to |
| 1022 |
<SI=0, BFR-NBR-C> is 0011 ("0001" OR'd with "0010"). |
| 1023 |
|
| 1024 |
The BIFT is used to map from the BFR-id of a BFER to the |
| 1025 |
corresponding F-BM and BFR-NBR. For example, Figure 3 shows the BIFT |
| 1026 |
that is derived from the BIRT of Figure 2. Note that BFR-ids 1 and 2 |
| 1027 |
have the same SI and the same BFR-NBR; hence, they have the |
| 1028 |
same F-BM. |
| 1029 |
|
| 1030 |
------------------------------------- |
| 1031 |
| BFR-id | F-BM | BFR-NBR | |
| 1032 |
| (SI:BitString) | | | |
| 1033 |
===================================== |
| 1034 |
| 1 (0:0001) | 0011 | C | |
| 1035 |
------------------------------------- |
| 1036 |
| 2 (0:0010) | 0011 | C | |
| 1037 |
------------------------------------- |
| 1038 |
| 3 (0:0100) | 0100 | E | |
| 1039 |
------------------------------------- |
| 1040 |
| 4 (0:1000) | 1000 | A | |
| 1041 |
------------------------------------- |
| 1042 |
|
| 1043 |
Figure 3: Bit Index Forwarding Table |
| 1044 |
|
| 1045 |
This BIFT is programmed into the data plane and used to forward |
| 1046 |
packets, applying the rules specified below in <a href="#section-6.5">Section 6.5</a>. |
| 1047 |
|
| 1048 |
|
| 1049 |
|
| 1050 |
|
| 1051 |
|
| 1052 |
|
| 1053 |
|
| 1054 |
|
| 1055 |
|
| 1056 |
|
| 1057 |
|
| 1058 |
|
| 1059 |
|
| 1060 |
<span class="grey">Wijnands, et al. Experimental [Page 19]</span></pre> |
| 1061 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-20" ></span> |
| 1062 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1063 |
|
| 1064 |
|
| 1065 |
<span class="h3"><a class="selflink" id="section-6.5" href="#section-6.5">6.5</a>. The BIER Forwarding Procedure</span> |
| 1066 |
|
| 1067 |
Below is the procedure that a BFR uses for forwarding a |
| 1068 |
BIER-encapsulated packet. |
| 1069 |
|
| 1070 |
1. Determine the packet's SI, BitStringLength, and sub-domain. |
| 1071 |
|
| 1072 |
2. If the BitString consists entirely of zeroes, discard the packet; |
| 1073 |
the forwarding process has been completed. Otherwise, proceed to |
| 1074 |
step 3. |
| 1075 |
|
| 1076 |
3. Find the position (call it "k") of the least significant (i.e., |
| 1077 |
of the rightmost) bit that is set in the packet's BitString. |
| 1078 |
(Remember, bits are numbered from 1, starting with the least |
| 1079 |
significant bit.) |
| 1080 |
|
| 1081 |
4. If bit k identifies the BFR itself, copy the packet, and send the |
| 1082 |
copy to the multicast flow overlay. Then clear bit k in the |
| 1083 |
original packet, and go to step 2. Otherwise, proceed to step 5. |
| 1084 |
|
| 1085 |
5. Use the value k, together with the SI, sub-domain, and |
| 1086 |
BitStringLength, as the "index" into the BIFT. |
| 1087 |
|
| 1088 |
6. Extract from the BIFT the F-BM and the BFR-NBR. |
| 1089 |
|
| 1090 |
7. Copy the packet. Update the copy's BitString by AND'ing it with |
| 1091 |
the F-BM (i.e., PacketCopy->BitString &= F-BM). Then forward the |
| 1092 |
copy to the BFR-NBR. (If the BFR-NBR is null, the copy is just |
| 1093 |
discarded.) Note that when a packet is forwarded to a particular |
| 1094 |
BFR-NBR, its BitString identifies only those BFERs that are to be |
| 1095 |
reached via that BFR-NBR. |
| 1096 |
|
| 1097 |
8. Now update the original packet's BitString by AND'ing it with the |
| 1098 |
INVERSE of the F-BM (i.e., Packet->BitString &= ~F-BM). (This |
| 1099 |
clears the bits that identify the BFERs to which a copy of the |
| 1100 |
packet has just been forwarded.) Go to step 2. |
| 1101 |
|
| 1102 |
This procedure causes the packet to be forwarded to a particular |
| 1103 |
BFR-NBR only once. The number of lookups in the BIFT is the same as |
| 1104 |
the number of BFR-NBRs to which the packet must be forwarded; it is |
| 1105 |
not necessary to do a separate lookup for each destination BFER. |
| 1106 |
|
| 1107 |
When a packet is sent to a particular BFR-NBR, the BitString is not |
| 1108 |
the only part of the BIER header that needs to be modified. If there |
| 1109 |
is a TTL field in the BIER header, it will need to be decremented. |
| 1110 |
In addition, when either of the encapsulations of [<a href="#ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>] |
| 1111 |
is used, the BIFT-id field is likely to require modification, based |
| 1112 |
on signaling from the BFR-NBR to which the packet is being sent. The |
| 1113 |
|
| 1114 |
|
| 1115 |
|
| 1116 |
<span class="grey">Wijnands, et al. Experimental [Page 20]</span></pre> |
| 1117 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-21" ></span> |
| 1118 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1119 |
|
| 1120 |
|
| 1121 |
BIFT-id field of an incoming BIER packet implicitly identifies an SI, |
| 1122 |
a sub-domain, and a BitStringLength. If the packet is sent to a |
| 1123 |
particular BFR-NBR, the BIFT-id field must be changed to whatever |
| 1124 |
value that BFR-NBR has advertised for the same SI, sub-domain, and |
| 1125 |
BitStringLength. (If the encapsulation of Section 2.1 of |
| 1126 |
[<a href="#ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>] is used, this is essentially an MPLS label swap |
| 1127 |
operation.) |
| 1128 |
|
| 1129 |
Suppose it has been decided (by the above rules) to send a packet to |
| 1130 |
a particular BFR-NBR. If that BFR-NBR is connected via multiple |
| 1131 |
parallel interfaces, it may be desirable to apply some form of load |
| 1132 |
balancing. Load-balancing algorithms are outside the scope of this |
| 1133 |
document. However, if the packet's encapsulation contains an entropy |
| 1134 |
field, the entropy field SHOULD be respected; two packets with the |
| 1135 |
same value of the entropy field SHOULD be sent on the same interface |
| 1136 |
(if possible). |
| 1137 |
|
| 1138 |
In some cases, the routing underlay may provide multiple equal-cost |
| 1139 |
paths (through different BFR-NBRs) to a given BFER. This is known as |
| 1140 |
"Equal-Cost Multipath" (ECMP). The procedures described in this |
| 1141 |
section must be augmented in order to support load balancing over |
| 1142 |
ECMP. The necessary augmentations can be found in <a href="#section-6.7">Section 6.7</a>. |
| 1143 |
|
| 1144 |
In the event that unicast traffic to the BFR-NBR is being sent via a |
| 1145 |
"bypass tunnel" of some sort, the BIER-encapsulated multicast traffic |
| 1146 |
sent to the BFR-NBR SHOULD also be sent via that tunnel. This allows |
| 1147 |
any existing "fast reroute" schemes to be applied to multicast |
| 1148 |
traffic as well as to unicast traffic. |
| 1149 |
|
| 1150 |
Some examples of these forwarding procedures can be found in |
| 1151 |
<a href="#section-6.6">Section 6.6</a>. |
| 1152 |
|
| 1153 |
|
| 1154 |
|
| 1155 |
|
| 1156 |
|
| 1157 |
|
| 1158 |
|
| 1159 |
|
| 1160 |
|
| 1161 |
|
| 1162 |
|
| 1163 |
|
| 1164 |
|
| 1165 |
|
| 1166 |
|
| 1167 |
|
| 1168 |
|
| 1169 |
|
| 1170 |
|
| 1171 |
|
| 1172 |
<span class="grey">Wijnands, et al. Experimental [Page 21]</span></pre> |
| 1173 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-22" ></span> |
| 1174 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1175 |
|
| 1176 |
|
| 1177 |
The rules given in this section can be represented by the following |
| 1178 |
pseudocode: |
| 1179 |
|
| 1180 |
void ForwardBitMaskPacket (Packet) |
| 1181 |
{ |
| 1182 |
SI=GetPacketSI(Packet); |
| 1183 |
Offset=SI*BitStringLength; |
| 1184 |
for (Index = GetFirstBitPosition(Packet->BitString); Index ; |
| 1185 |
Index = GetNextBitPosition(Packet->BitString, Index)) { |
| 1186 |
F-BM = BIFT[Index+Offset]->F-BM; |
| 1187 |
if (!F-BM) continue; |
| 1188 |
BFR-NBR = BIFT[Index+Offset]->BFR-NBR; |
| 1189 |
PacketCopy = Copy(Packet); |
| 1190 |
PacketCopy->BitString &= F-BM; |
| 1191 |
PacketSend(PacketCopy, BFR-NBR); |
| 1192 |
Packet->BitString &= ~F-BM; |
| 1193 |
} |
| 1194 |
} |
| 1195 |
|
| 1196 |
Figure 4: Pseudocode |
| 1197 |
|
| 1198 |
This pseudocode assumes that, at a given BFER, the BFR-NBR entry |
| 1199 |
corresponding to the BFER's own BFR-id will be the BFER's own |
| 1200 |
BFR-prefix. It also assumes that the corresponding F-BM has only |
| 1201 |
one bit set, the bit representing the BFER itself. In this case, the |
| 1202 |
"PacketSend" function sends the packet to the multicast flow overlay. |
| 1203 |
|
| 1204 |
This pseudocode also assumes that the F-BM for the null next hop |
| 1205 |
contains a 1 in a given bit position if and only if that bit position |
| 1206 |
corresponds to either an unused BFR-id or an unreachable BFER. When |
| 1207 |
the BFR-NBR is null, the "PacketSend" function discards the packet. |
| 1208 |
|
| 1209 |
|
| 1210 |
|
| 1211 |
|
| 1212 |
|
| 1213 |
|
| 1214 |
|
| 1215 |
|
| 1216 |
|
| 1217 |
|
| 1218 |
|
| 1219 |
|
| 1220 |
|
| 1221 |
|
| 1222 |
|
| 1223 |
|
| 1224 |
|
| 1225 |
|
| 1226 |
|
| 1227 |
|
| 1228 |
<span class="grey">Wijnands, et al. Experimental [Page 22]</span></pre> |
| 1229 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-23" ></span> |
| 1230 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1231 |
|
| 1232 |
|
| 1233 |
<span class="h3"><a class="selflink" id="section-6.6" href="#section-6.6">6.6</a>. Examples of BIER Forwarding</span> |
| 1234 |
|
| 1235 |
In this section, we give two examples of BIER forwarding, based on |
| 1236 |
the topology in Figure 1. In these examples, all packets have been |
| 1237 |
assigned to the default sub-domain, all packets have SI=0, and the |
| 1238 |
BitStringLength is 4. Figure 5 shows the BIFT entries for SI=0 only. |
| 1239 |
For compactness, we show the first column of the BIFT, the BFR-id, |
| 1240 |
only as an integer. |
| 1241 |
|
| 1242 |
BFR-A BIFT BFR-B BIFT BFR-C BIFT |
| 1243 |
------------------- ------------------- ------------------- |
| 1244 |
| Id | F-BM | NBR | | Id | F-BM | NBR | | Id | F-BM | NBR | |
| 1245 |
=================== =================== =================== |
| 1246 |
| 1 | 0111 | B | | 1 | 0011 | C | | 1 | 0001 | D | |
| 1247 |
------------------- ------------------- ------------------- |
| 1248 |
| 2 | 0111 | B | | 2 | 0011 | C | | 2 | 0010 | F | |
| 1249 |
------------------- ------------------- ------------------- |
| 1250 |
| 3 | 0111 | B | | 3 | 0100 | E | | 3 | 1100 | B | |
| 1251 |
------------------- ------------------- ------------------- |
| 1252 |
| 4 | 1000 | A | | 4 | 1000 | A | | 4 | 1100 | B | |
| 1253 |
------------------- ------------------- ------------------- |
| 1254 |
|
| 1255 |
Figure 5: BIFTs Used in the Forwarding Examples |
| 1256 |
|
| 1257 |
<span class="h4"><a class="selflink" id="section-6.6.1" href="#section-6.6.1">6.6.1</a>. Example 1</span> |
| 1258 |
|
| 1259 |
BFR-D, BFR-E, and BFR-F are BFERs. BFR-A is the BFIR. Suppose that |
| 1260 |
BFIR-A has learned from the multicast flow overlay that BFER-D is |
| 1261 |
interested in a given multicast flow. If BFIR-A receives a packet of |
| 1262 |
that flow from outside the BIER domain, BFIR-A applies the BIER |
| 1263 |
encapsulation to the packet. The encapsulation must be such that the |
| 1264 |
SI is zero. The encapsulation also includes a BitString, with just |
| 1265 |
bit 1 set and with all other bits clear (i.e., 0001). This indicates |
| 1266 |
that BFER-D is the only BFER that needs to receive the packet. |
| 1267 |
BFIR-A then follows the procedures of <a href="#section-6.5">Section 6.5</a>, as follows: |
| 1268 |
|
| 1269 |
o Since the packet's BitString is 0001, BFIR-A finds that the first |
| 1270 |
bit in the string is bit 1. Looking at entry 1 in its BIFT, BFR-A |
| 1271 |
determines that the bit mask F-BM is 0111 and the BFR-NBR is |
| 1272 |
BFR-B. |
| 1273 |
|
| 1274 |
o BFR-A then makes a copy of the packet and applies the F-BM to the |
| 1275 |
copy: Copy->BitString &= 0111. The copy's BitString is now 0001 |
| 1276 |
(0001 & 0111). |
| 1277 |
|
| 1278 |
o The copy is now sent to BFR-B. |
| 1279 |
|
| 1280 |
|
| 1281 |
|
| 1282 |
|
| 1283 |
|
| 1284 |
<span class="grey">Wijnands, et al. Experimental [Page 23]</span></pre> |
| 1285 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-24" ></span> |
| 1286 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1287 |
|
| 1288 |
|
| 1289 |
o BFR-A then updates the packet's BitString by applying the inverse |
| 1290 |
of the F-BM: Packet->BitString &= ~F-BM. As a result, the |
| 1291 |
packet's BitString is now 0000 (0001 & 1000). |
| 1292 |
|
| 1293 |
o As the packet's BitString is now zero, the forwarding procedure is |
| 1294 |
complete. |
| 1295 |
|
| 1296 |
When BFR-B receives the multicast packet from BFR-A, it follows the |
| 1297 |
same procedure. The result is that a copy of the packet, with a |
| 1298 |
BitString of 0001, is sent to BFR-C. BFR-C applies the same |
| 1299 |
procedures and, as a result, sends a copy of the packet, with a |
| 1300 |
BitString of 0001, to BFR-D. |
| 1301 |
|
| 1302 |
At BFER-D, the BIFT entry (not pictured) for BFR-id 1 will specify an |
| 1303 |
F-BM of 0001 and a BFR-NBR of BFR-D itself. This will cause a copy |
| 1304 |
of the packet to be delivered to the multicast flow overlay at BFR-D. |
| 1305 |
The packet's BitString will be set to 0000, and the packet will not |
| 1306 |
be forwarded any further. |
| 1307 |
|
| 1308 |
<span class="h4"><a class="selflink" id="section-6.6.2" href="#section-6.6.2">6.6.2</a>. Example 2</span> |
| 1309 |
|
| 1310 |
This example is similar to example 1, except that BFIR-A has learned |
| 1311 |
from the multicast flow overlay that both BFER-D and BFER-E are |
| 1312 |
interested in a given multicast flow. If BFIR-A receives a packet of |
| 1313 |
that flow from outside the BIER domain, BFIR-A applies the BIER |
| 1314 |
encapsulation to the packet. The encapsulation must be such that the |
| 1315 |
SI is zero. The encapsulation also includes a BitString with |
| 1316 |
two bits set: bit 1 is set (as in example 1) to indicate that BFR-D |
| 1317 |
is a BFER for this packet, and bit 3 is set to indicate that BFR-E is |
| 1318 |
a BFER for this packet. That is, the BitString (assuming again a |
| 1319 |
BitStringLength of 4) is 0101. To forward the packet, BFIR-A follows |
| 1320 |
the procedures of <a href="#section-6.5">Section 6.5</a>, as follows: |
| 1321 |
|
| 1322 |
o Since the packet's BitString is 0101, BFIR-A finds that the first |
| 1323 |
bit in the string is bit 1. Looking at entry 1 in its BIFT, BFR-A |
| 1324 |
determines that the bit mask F-BM is 0111 and the BFR-NBR is |
| 1325 |
BFR-B. |
| 1326 |
|
| 1327 |
o BFR-A then makes a copy of the packet and applies the F-BM to the |
| 1328 |
copy: Copy->BitString &= 0111. The copy's BitString is now 0101 |
| 1329 |
(0101 & 0111). |
| 1330 |
|
| 1331 |
o The copy is now sent to BFR-B. |
| 1332 |
|
| 1333 |
|
| 1334 |
|
| 1335 |
|
| 1336 |
|
| 1337 |
|
| 1338 |
|
| 1339 |
|
| 1340 |
<span class="grey">Wijnands, et al. Experimental [Page 24]</span></pre> |
| 1341 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-25" ></span> |
| 1342 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1343 |
|
| 1344 |
|
| 1345 |
o BFR-A then updates the packet's BitString by applying the inverse |
| 1346 |
of the F-BM: Packet->BitString &= ~F-BM. As a result, the |
| 1347 |
packet's BitString is now 0000 (0101 & 1000). |
| 1348 |
|
| 1349 |
o As the packet's BitString is now zero, the forwarding procedure is |
| 1350 |
complete. |
| 1351 |
|
| 1352 |
When BFR-B receives the multicast packet from BFR-A, it follows the |
| 1353 |
procedure of <a href="#section-6.5">Section 6.5</a>, as follows: |
| 1354 |
|
| 1355 |
o Since the packet's BitString is 0101, BFR-B finds that the first |
| 1356 |
bit in the string is bit 1. Looking at entry 1 in its BIFT, BFR-B |
| 1357 |
determines that the bit mask F-BM is 0011 and the BFR-NBR is |
| 1358 |
BFR-C. |
| 1359 |
|
| 1360 |
o BFR-B then makes a copy of the packet and applies the F-BM to the |
| 1361 |
copy: Copy->BitString &= 0011. The copy's BitString is now 0001 |
| 1362 |
(0101 & 0011). |
| 1363 |
|
| 1364 |
o The copy is now sent to BFR-C. |
| 1365 |
|
| 1366 |
o BFR-B then updates the packet's BitString by applying the inverse |
| 1367 |
of the F-BM: Packet->BitString &= ~F-BM. As a result, the |
| 1368 |
packet's BitString is now 0100 (0101 & 1100). |
| 1369 |
|
| 1370 |
o Now BFR-B finds the next bit in the packet's (modified) BitString. |
| 1371 |
This is bit 3. Looking at entry 3 in its BIFT, BFR-B determines |
| 1372 |
that the F-BM is 0100 and the BFR-NBR is BFR-E. |
| 1373 |
|
| 1374 |
o BFR-B then makes a copy of the packet and applies the F-BM to the |
| 1375 |
copy: Copy->BitString &= 0100. The copy's BitString is now 0100 |
| 1376 |
(0100 & 0100). |
| 1377 |
|
| 1378 |
o The copy is now sent to BFR-E. |
| 1379 |
|
| 1380 |
o BFR-B then updates the packet's BitString by applying the inverse |
| 1381 |
of the F-BM: Packet->BitString &= ~F-BM. As a result, the |
| 1382 |
packet's BitString is now 0000 (0100 & 1011). |
| 1383 |
|
| 1384 |
o As the packet's BitString is now zero, the forwarding procedure is |
| 1385 |
complete. |
| 1386 |
|
| 1387 |
Thus, BFR-B forwards two copies of the packet. One copy of the |
| 1388 |
packet, with BitString 0001, has now been sent from BFR-B to BFR-C. |
| 1389 |
Following the same procedures, BFR-C will forward the packet to |
| 1390 |
BFER-D. |
| 1391 |
|
| 1392 |
|
| 1393 |
|
| 1394 |
|
| 1395 |
|
| 1396 |
<span class="grey">Wijnands, et al. Experimental [Page 25]</span></pre> |
| 1397 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-26" ></span> |
| 1398 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1399 |
|
| 1400 |
|
| 1401 |
At BFER-D, the BIFT entry (not pictured) for BFR-id 1 will specify an |
| 1402 |
F-BM of 0001 and a BFR-NBR of BFR-D itself. This will cause a copy |
| 1403 |
of the packet to be delivered to the multicast flow overlay at BFR-D. |
| 1404 |
The packet's BitString will be set to 0000, and the packet will not |
| 1405 |
be forwarded any further. |
| 1406 |
|
| 1407 |
The other copy of the packet has been sent from BFR-B to BFER-E, with |
| 1408 |
BitString 0100. |
| 1409 |
|
| 1410 |
At BFER-E, the BIFT entry (not pictured) for BFR-id 3 will specify an |
| 1411 |
F-BM of 0100 and a BFR-NBR of BFR-E itself. This will cause a copy |
| 1412 |
of the packet to be delivered to the multicast flow overlay at BFR-E. |
| 1413 |
The packet's BitString will be set to 0000, and the packet will not |
| 1414 |
be forwarded any further. |
| 1415 |
|
| 1416 |
<span class="h3"><a class="selflink" id="section-6.7" href="#section-6.7">6.7</a>. Equal-Cost Multipath Forwarding</span> |
| 1417 |
|
| 1418 |
In many networks, the routing underlay will provide multiple |
| 1419 |
equal-cost paths from a given BFR to a given BFER. When forwarding |
| 1420 |
multicast packets through the network, it can be beneficial to take |
| 1421 |
advantage of this by load-balancing among those paths. This feature |
| 1422 |
is known as "Equal-Cost Multipath (ECMP) forwarding". |
| 1423 |
|
| 1424 |
BIER supports ECMP forwarding, but the procedures of <a href="#section-6.5">Section 6.5</a> must |
| 1425 |
be modified slightly. Two ECMP procedures are defined. In the first |
| 1426 |
(described in <a href="#section-6.7.1">Section 6.7.1</a>), the choice among equal-cost paths taken |
| 1427 |
by a given packet from a given BFR to a given BFER depends on |
| 1428 |
(a) routing, (b) the packet's entropy, and (c) the other BFERs to |
| 1429 |
which that packet is destined. In the second (described in |
| 1430 |
<a href="#section-6.7.2">Section 6.7.2</a>), the choice depends only upon the packet's entropy. |
| 1431 |
|
| 1432 |
There are trade-offs between the two forwarding procedures described |
| 1433 |
here. In the procedure of <a href="#section-6.7.1">Section 6.7.1</a>, the number of packet |
| 1434 |
replications is minimized. The procedure in <a href="#section-6.7.1">Section 6.7.1</a> also uses |
| 1435 |
less memory in the BFR. In the procedure of <a href="#section-6.7.2">Section 6.7.2</a>, the path |
| 1436 |
traveled by a given packet from a given BFR to a given BFER is |
| 1437 |
independent of the other BFERs to which the packet is destined. |
| 1438 |
While the procedures of <a href="#section-6.7.2">Section 6.7.2</a> may cause more replications, |
| 1439 |
they provide a more predictable behavior. |
| 1440 |
|
| 1441 |
The two procedures described here operate on identical packet formats |
| 1442 |
and will interoperate correctly. However, if deterministic behavior |
| 1443 |
is desired, then all BFRs would need to use the procedure from |
| 1444 |
<a href="#section-6.7.2">Section 6.7.2</a>. |
| 1445 |
|
| 1446 |
|
| 1447 |
|
| 1448 |
|
| 1449 |
|
| 1450 |
|
| 1451 |
|
| 1452 |
<span class="grey">Wijnands, et al. Experimental [Page 26]</span></pre> |
| 1453 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-27" ></span> |
| 1454 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1455 |
|
| 1456 |
|
| 1457 |
<span class="h4"><a class="selflink" id="section-6.7.1" href="#section-6.7.1">6.7.1</a>. Non-deterministic ECMP</span> |
| 1458 |
|
| 1459 |
Figure 6 shows the operation of non-deterministic ECMP in BIER. |
| 1460 |
|
| 1461 |
BFR-A BIFT BFR-B BIFT BFR-C BIFT |
| 1462 |
------------------- ------------------- ------------------- |
| 1463 |
| Id | F-BM | NBR | | Id | F-BM | NBR | | Id | F-BM | NBR | |
| 1464 |
=================== =================== =================== |
| 1465 |
| 1 | 0111 | B | | 1 | 0011 | C | | 1 | 0001 | D | |
| 1466 |
------------------- ------------------- ------------------- |
| 1467 |
| 2 | 0111 | B | | 2 | 0011 | C | | 2 | 0010 | F | |
| 1468 |
------------------- | | 0110 | E | ------------------- |
| 1469 |
| 3 | 0111 | B | ------------------- | 3 | 1100 | B | |
| 1470 |
------------------- | 3 | 0110 | E | ------------------- |
| 1471 |
| 4 | 1000 | A | ------------------| | 4 | 1100 | B | |
| 1472 |
------------------- | 4 | 1000 | A | ------------------- |
| 1473 |
------------------- |
| 1474 |
|
| 1475 |
( A ) ------------ ( B ) ------------ ( C ) ------------ ( D ) |
| 1476 |
4 (0:1000) \ \ 1 (0:0001) |
| 1477 |
\ \ |
| 1478 |
( E ) ------------ ( F ) |
| 1479 |
3 (0:0100) 2 (0:0010) |
| 1480 |
|
| 1481 |
Figure 6: Example of Non-deterministic ECMP |
| 1482 |
|
| 1483 |
In this example, BFR-B has two equal-cost paths to reach BFER-F: one |
| 1484 |
via BFR-C and one via BFR-E. Since the BFR-id of BFER-F is 2, this |
| 1485 |
is reflected in entry 2 of BFR-B's BIFT. Entry 2 shows that BFR-B |
| 1486 |
has a choice of two BFR-NBRs for BFER-B and that a different F-BM is |
| 1487 |
associated with each choice. When BFR-B looks up entry 2 in the |
| 1488 |
BIFT, it can choose either BFR-NBR. However, when following the |
| 1489 |
procedures of <a href="#section-6.5">Section 6.5</a>, it MUST use the F-BM corresponding to the |
| 1490 |
BFR-NBR that it chooses. |
| 1491 |
|
| 1492 |
How the choice is made is an implementation matter. However, the |
| 1493 |
usual rules for ECMP apply: packets of a given flow SHOULD NOT be |
| 1494 |
split among two paths, and any entropy field in the packet's |
| 1495 |
encapsulation SHOULD be respected. |
| 1496 |
|
| 1497 |
Note, however, that by the rules of <a href="#section-6.5">Section 6.5</a>, any packet destined |
| 1498 |
for both BFER-D and BFER-F will be sent via BFR-C. |
| 1499 |
|
| 1500 |
|
| 1501 |
|
| 1502 |
|
| 1503 |
|
| 1504 |
|
| 1505 |
|
| 1506 |
|
| 1507 |
|
| 1508 |
<span class="grey">Wijnands, et al. Experimental [Page 27]</span></pre> |
| 1509 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-28" ></span> |
| 1510 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1511 |
|
| 1512 |
|
| 1513 |
<span class="h4"><a class="selflink" id="section-6.7.2" href="#section-6.7.2">6.7.2</a>. Deterministic ECMP</span> |
| 1514 |
|
| 1515 |
With the procedures of <a href="#section-6.7.1">Section 6.7.1</a>, where ECMP paths exist, the |
| 1516 |
path a packet takes to reach any particular BFER depends not only on |
| 1517 |
routing and on the packet's entropy but also on the set of other |
| 1518 |
BFERs to which the packet is destined. |
| 1519 |
|
| 1520 |
For example, consider the following scenario in the network of |
| 1521 |
Figure 6. |
| 1522 |
|
| 1523 |
o There is a sequence of packets being transmitted by BFR-A, some of |
| 1524 |
which are destined for both D and F and some of which are destined |
| 1525 |
only for F. |
| 1526 |
|
| 1527 |
o All the packets in this sequence have the same entropy value (call |
| 1528 |
it "Q"). |
| 1529 |
|
| 1530 |
o At BFR-B, when a packet with entropy value Q is forwarded via |
| 1531 |
entry 2 in the BIFT, the packet is sent to E. |
| 1532 |
|
| 1533 |
Using the forwarding procedure of <a href="#section-6.7.1">Section 6.7.1</a>, packets of this |
| 1534 |
sequence that are destined for both D and F are forwarded according |
| 1535 |
to entry 1 in the BIFT and thus will reach F via the path A-B-C-F. |
| 1536 |
However, packets of this sequence that are destined only for F are |
| 1537 |
forwarded according to entry 2 in the BIFT and thus will reach F via |
| 1538 |
the path A-B-E-F. |
| 1539 |
|
| 1540 |
That procedure minimizes the number of packets transmitted by BFR-B. |
| 1541 |
However, consider the following scenario: |
| 1542 |
|
| 1543 |
o Beginning at time t0, the multicast flow in question needs to be |
| 1544 |
received ONLY by BFER-F. |
| 1545 |
|
| 1546 |
o Beginning at a later time, t1, the flow needs to be received by |
| 1547 |
both BFER-D and BFER-F. |
| 1548 |
|
| 1549 |
o Beginning at a later time, t2, the flow no longer needs to be |
| 1550 |
received by D, but still needs to be received by F. |
| 1551 |
|
| 1552 |
Then, from t0 until t1, the flow will travel to F via the path |
| 1553 |
A-B-E-F. From t1 until t2, the flow will travel to F via the path |
| 1554 |
A-B-C-F. And from t2, the flow will again travel to F via the path |
| 1555 |
A-B-E-F. |
| 1556 |
|
| 1557 |
The problem is that if D repeatedly joins and leaves the flow, the |
| 1558 |
flow's path from B to F will keep switching. This could cause F to |
| 1559 |
receive packets out of order. It also makes troubleshooting |
| 1560 |
difficult. For example, if there is some problem on the E-F link, |
| 1561 |
|
| 1562 |
|
| 1563 |
|
| 1564 |
<span class="grey">Wijnands, et al. Experimental [Page 28]</span></pre> |
| 1565 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-29" ></span> |
| 1566 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1567 |
|
| 1568 |
|
| 1569 |
receivers at F will get good service when the flow is also going to D |
| 1570 |
(avoiding the E-F link) but bad service when the flow is not going |
| 1571 |
to D. Since it is hard to know which path is being used at any given |
| 1572 |
time, this may be hard to troubleshoot. Also, it is very difficult |
| 1573 |
to perform a traceroute that is known to follow the path taken by the |
| 1574 |
flow at any given time. |
| 1575 |
|
| 1576 |
The source of this difficulty is that, in the procedures of |
| 1577 |
<a href="#section-6.7.1">Section 6.7.1</a>, the path taken by a particular flow to a particular |
| 1578 |
BFER depends upon whether there are lower-numbered BFERs that are |
| 1579 |
also receiving the flow. Thus, the choice among the ECMP paths is |
| 1580 |
fundamentally non-deterministic. |
| 1581 |
|
| 1582 |
Deterministic forwarding can be achieved by using multiple BIFTs, |
| 1583 |
such that each row in a BIFT has only one path to each destination |
| 1584 |
but the multiple ECMP paths to any particular destination are spread |
| 1585 |
across the multiple tables. When a BIER-encapsulated packet arrives |
| 1586 |
to be forwarded, the BFR uses a hash of the BIER entropy field to |
| 1587 |
determine which BIFT to use, and then the normal BIER forwarding |
| 1588 |
algorithm (as described in Sections <a href="#section-6.5">6.5</a> and <a href="#section-6.6">6.6</a>) is used with the |
| 1589 |
selected BIFT. |
| 1590 |
|
| 1591 |
As an example, suppose there are two paths to destination X (call |
| 1592 |
them "X1" and "X2") and four paths to destination Y (call them "Y1", |
| 1593 |
"Y2", "Y3", and "Y4"). If there are, say, four BIFTs, one BIFT would |
| 1594 |
have paths X1 and Y1, one would have X1 and Y2, one would have X2 and |
| 1595 |
Y3, and one would have X2 and Y4. If traffic to X is split evenly |
| 1596 |
among these four BIFTs, the traffic will be split evenly between the |
| 1597 |
two paths to X; if traffic to Y is split evenly among these four |
| 1598 |
BIFTs, the traffic will be split evenly between the four paths to Y. |
| 1599 |
|
| 1600 |
Note that if there are three paths to one destination and four paths |
| 1601 |
to another, 12 BIFTs would be required in order to get even splitting |
| 1602 |
of the load to each of those two destinations. Of course, each BIFT |
| 1603 |
uses some memory, and one might be willing to have less optimal |
| 1604 |
splitting in order to have fewer BIFTs. How that trade-off is made |
| 1605 |
is an implementation or deployment decision. |
| 1606 |
|
| 1607 |
<span class="h3"><a class="selflink" id="section-6.8" href="#section-6.8">6.8</a>. Prevention of Loops and Duplicates</span> |
| 1608 |
|
| 1609 |
The BitString in a BIER-encapsulated packet specifies the set of |
| 1610 |
BFERs to which that packet is to be forwarded. When a |
| 1611 |
BIER-encapsulated packet is replicated, no two copies of the packet |
| 1612 |
will ever have a BFER in common. If one of the packet's BFERs |
| 1613 |
forwards the packet further, that BFER will first clear the bit that |
| 1614 |
identifies itself. As a result, duplicate delivery of packets is not |
| 1615 |
possible with BIER. |
| 1616 |
|
| 1617 |
|
| 1618 |
|
| 1619 |
|
| 1620 |
<span class="grey">Wijnands, et al. Experimental [Page 29]</span></pre> |
| 1621 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-30" ></span> |
| 1622 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1623 |
|
| 1624 |
|
| 1625 |
As long as the routing underlay provides a loop-free path between |
| 1626 |
each pair of BFRs, BIER-encapsulated packets will not loop. Since |
| 1627 |
the BIER layer does not create any paths of its own, there is no need |
| 1628 |
for any BIER-specific loop-prevention techniques beyond the |
| 1629 |
forwarding procedures specified in <a href="#section-6.5">Section 6.5</a>. |
| 1630 |
|
| 1631 |
If, at some time, the routing underlay is not providing a loop-free |
| 1632 |
path between BFIR-A and BFER-B, then BIER-encapsulated packets may |
| 1633 |
loop while traveling from BFIR-A to BFER-B. However, such loops will |
| 1634 |
never result in delivery of duplicate packets to BFER-B. |
| 1635 |
|
| 1636 |
These properties of BIER eliminate the need for the "Reverse Path |
| 1637 |
Forwarding" (RPF) check that is used in conventional IP multicast |
| 1638 |
forwarding. |
| 1639 |
|
| 1640 |
<span class="h3"><a class="selflink" id="section-6.9" href="#section-6.9">6.9</a>. When Some Nodes Do Not Support BIER</span> |
| 1641 |
|
| 1642 |
The procedures of <a href="#section-6.2">Section 6.2</a> presuppose that, within a given BIER |
| 1643 |
domain, all the nodes adjacent to a given BFR in a given routing |
| 1644 |
underlay are also BFRs. However, it is possible to use BIER even |
| 1645 |
when this is not the case, as long as the ingress and egress nodes |
| 1646 |
are BFRs. In this section, we describe procedures that can be used |
| 1647 |
if the routing underlay is an SPF-based IGP that computes a |
| 1648 |
shortest-path tree from each node to all other nodes in the domain. |
| 1649 |
|
| 1650 |
At a given BFR, say "BFR-B", start with a copy of the IGP-computed |
| 1651 |
shortest-path tree from BFR-B to each router in the domain. (This |
| 1652 |
tree is computed by the SPF algorithm of the IGP.) Let's call this |
| 1653 |
copy the "BIER-SPF tree rooted at BFR-B". BFR-B then modifies this |
| 1654 |
BIER-SPF tree as follows. |
| 1655 |
|
| 1656 |
1. BFR-B looks in turn at each of its child nodes on the BIER-SPF |
| 1657 |
tree. |
| 1658 |
|
| 1659 |
2. If one of the child nodes does not support BIER, BFR-B removes |
| 1660 |
that node from the tree. The child nodes of the node that has |
| 1661 |
just been removed are then re-parented on the tree, so that BFR-B |
| 1662 |
now becomes their parent. |
| 1663 |
|
| 1664 |
3. BFR-B then continues to look at each of its child nodes, |
| 1665 |
including any nodes that have been re-parented to BFR-B as a |
| 1666 |
result of the previous step. |
| 1667 |
|
| 1668 |
When all of the child nodes (the original child nodes plus any new |
| 1669 |
ones) have been examined, BFR-B's children on the BIER-SPF tree will |
| 1670 |
all be BFRs. |
| 1671 |
|
| 1672 |
|
| 1673 |
|
| 1674 |
|
| 1675 |
|
| 1676 |
<span class="grey">Wijnands, et al. Experimental [Page 30]</span></pre> |
| 1677 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-31" ></span> |
| 1678 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1679 |
|
| 1680 |
|
| 1681 |
When the BIFT is constructed, BFR-B's child nodes on the BIER-SPF |
| 1682 |
tree are considered to be the BFR-NBRs. The F-BMs must be computed |
| 1683 |
appropriately, based on the BFR-NBRs. |
| 1684 |
|
| 1685 |
BFR-B may now have BFR-NBRs that are not "directly connected" to |
| 1686 |
BFR-B via Layer 2. To send a packet to one of these BFR-NBRs, BFR-B |
| 1687 |
will have to send the packet through a unicast tunnel. In an MPLS |
| 1688 |
network, this may be as simple as finding the IGP unicast next hop to |
| 1689 |
the child node and pushing on (above the BIER encapsulation header) |
| 1690 |
an MPLS label that the IGP next hop has bound to an address of the |
| 1691 |
child node. (This assumes that the packet is using an MPLS-based |
| 1692 |
BIER encapsulation, such as the one specified in Section 2.1 of |
| 1693 |
[<a href="#ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>].) Of course, the BIFT-id in the BIER |
| 1694 |
encapsulation header must be the BIFT-id advertised by the child node |
| 1695 |
for the packet's SI, sub-domain, and BitStringLength. |
| 1696 |
|
| 1697 |
If for some reason the unicast tunnel cannot be an MPLS tunnel, any |
| 1698 |
other kind of tunnel can be used, as long as the encapsulation for |
| 1699 |
that tunnel type has a way of indicating that the payload is a |
| 1700 |
BIER-encapsulated packet. |
| 1701 |
|
| 1702 |
Note that if a BIER-encapsulated packet is not using an MPLS-based |
| 1703 |
BIER encapsulation, it will not be possible to send it through an |
| 1704 |
MPLS tunnel unless it is known that the tunnel only carries BIER |
| 1705 |
packets; this is because MPLS has no "next protocol type" field. |
| 1706 |
This is not a problem if an MPLS-based BIER encapsulation is used, |
| 1707 |
because in that case the BIER encapsulation begins with an MPLS label |
| 1708 |
that identifies the packet as a BIER-encapsulated packet. |
| 1709 |
|
| 1710 |
Of course, the above is not meant as an implementation technique, |
| 1711 |
just as a functional description. |
| 1712 |
|
| 1713 |
While the above description assumes that the routing underlay |
| 1714 |
provides an SPF tree, it may also be applicable to other types of |
| 1715 |
routing underlays. |
| 1716 |
|
| 1717 |
The technique above can also be used to provide "node protection" |
| 1718 |
(i.e., to provide fast reroute around nodes that are believed to have |
| 1719 |
failed). If BFR-B has a failed BFR-NBR, BFR-B can remove the failed |
| 1720 |
BFR-NBR from the BIER-SPF tree and can then re-parent the child |
| 1721 |
BFR-NBRs of the failed BFR-NBR so that they appear to be BFR-B's own |
| 1722 |
child nodes on the tree (i.e., so that they appear to be BFR-B's |
| 1723 |
BFR-NBRs). The usual BIER forwarding procedures then apply. |
| 1724 |
However, getting the packet from BFR-B to the child nodes of the |
| 1725 |
failed BFR-NBR is a bit more complicated, as it may require using a |
| 1726 |
unicast bypass tunnel to get around the failed node. |
| 1727 |
|
| 1728 |
|
| 1729 |
|
| 1730 |
|
| 1731 |
|
| 1732 |
<span class="grey">Wijnands, et al. Experimental [Page 31]</span></pre> |
| 1733 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-32" ></span> |
| 1734 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1735 |
|
| 1736 |
|
| 1737 |
A simpler variant of step 2 above would be the following: |
| 1738 |
|
| 1739 |
If one of the child nodes does not support BIER, BFR-B removes |
| 1740 |
that node from the tree. All BFERs that are reached through that |
| 1741 |
child node are then re-parented on the tree, so that BFR-B now |
| 1742 |
becomes their parent. |
| 1743 |
|
| 1744 |
This variant is simpler because the set of BFERs that are reached |
| 1745 |
through a particular child node of BFR-B can be determined from the |
| 1746 |
F-BM in the BIFT. However, if this variant is used, the results are |
| 1747 |
less optimal, because packets will be unicast directly from BFR-B to |
| 1748 |
the BFERs that are reachable through the non-BIER child node. |
| 1749 |
|
| 1750 |
When using a unicast MPLS tunnel to get a packet to a BFR-NBR: |
| 1751 |
|
| 1752 |
o The TTL of the MPLS label entry representing the tunnel SHOULD be |
| 1753 |
set to a large value, rather than being copied from the TTL value |
| 1754 |
from the BIER encapsulation header, and |
| 1755 |
|
| 1756 |
o When the tunnel labels are popped off, the TTL from the tunnel |
| 1757 |
labels SHOULD NOT be copied to the BIER encapsulation header. |
| 1758 |
|
| 1759 |
In other words, the TTL processing for the tunnel SHOULD be as |
| 1760 |
specified in [<a href="./rfc3443" title=""Time To Live (TTL) Processing in Multi-Protocol Label Switching (MPLS) Networks"">RFC3443</a>] for "Pipe Model" and "Short Pipe Model" Label |
| 1761 |
Switched Paths (LSPs). The same principle applies if the tunnels are |
| 1762 |
not MPLS tunnels; the BIER packet SHOULD NOT inherit the TTL from the |
| 1763 |
tunnel encapsulation. That way, the TTL of the BIER encapsulation |
| 1764 |
header constrains only the number of BFRs that the packet may |
| 1765 |
traverse, not the total number of hops. |
| 1766 |
|
| 1767 |
If two BIER packets have the same value in the entropy field of their |
| 1768 |
respective BIER headers and if both are transmitted through a given |
| 1769 |
tunnel, it is desirable for the tunnel encapsulation to preserve the |
| 1770 |
fact that the two packets have the same entropy. |
| 1771 |
|
| 1772 |
The material in this section presupposes that if a given router is a |
| 1773 |
BFR, then it supports BIER on all its interfaces. It is, however, |
| 1774 |
possible that a router will have some line cards that support BIER |
| 1775 |
and some that do not. In such a case, one can think of the router as |
| 1776 |
a "partial BFR" that supports BIER only on some of its interfaces. |
| 1777 |
If it is desired to deploy such partial BFRs, one can use the |
| 1778 |
multi-topology features of the IGP to set up a BIER-specific |
| 1779 |
topology. This topology would exclude all the non-BIER-capable |
| 1780 |
interfaces that attach to BFRs. BIER would then have to be run in a |
| 1781 |
sub-domain that is bound to this topology. If unicast tunnels are |
| 1782 |
used to bypass non-BFRs, either (a) the tunnels have to be restricted |
| 1783 |
to this topology or (b) the tunnel endpoints have to be BFRs that do |
| 1784 |
not have any non-BIER-capable interfaces. |
| 1785 |
|
| 1786 |
|
| 1787 |
|
| 1788 |
<span class="grey">Wijnands, et al. Experimental [Page 32]</span></pre> |
| 1789 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-33" ></span> |
| 1790 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1791 |
|
| 1792 |
|
| 1793 |
<span class="h3"><a class="selflink" id="section-6.10" href="#section-6.10">6.10</a>. Use of Different BitStringLengths within a Domain</span> |
| 1794 |
|
| 1795 |
The procedures of this section apply only when the same encapsulation |
| 1796 |
is used throughout the BIER domain. Consideration of the scenario |
| 1797 |
where both multiple encapsulations and multiple BitStringLengths are |
| 1798 |
used in a given BIER domain is outside the scope of this document. |
| 1799 |
|
| 1800 |
It is possible for different BFRs within a BIER domain to be using |
| 1801 |
different Imposition and/or Disposition BitStringLengths. As stated |
| 1802 |
in <a href="#section-3">Section 3</a>: |
| 1803 |
|
| 1804 |
"if a particular BFIR is provisioned to use a particular Imposition |
| 1805 |
BitStringLength and a particular Imposition sub-domain when imposing |
| 1806 |
the encapsulation on a given set of packets, all other BFRs with |
| 1807 |
BFR-ids in that sub-domain SHOULD be provisioned to process received |
| 1808 |
BIER packets with that BitStringLength (i.e., all other BFRs with |
| 1809 |
BFR-ids in that sub-domain SHOULD be provisioned with that |
| 1810 |
BitStringLength as a Disposition BitStringLength for that |
| 1811 |
sub-domain)." |
| 1812 |
|
| 1813 |
Note that mis-provisioning can result in "black holes". If a BFIR |
| 1814 |
creates a BIER packet with a particular BitStringLength and if that |
| 1815 |
packet needs to travel through a BFR that cannot process received |
| 1816 |
BIER packets with that BitStringLength, then it may be impossible to |
| 1817 |
forward the packet to all of the BFERs identified in its BIER header. |
| 1818 |
<a href="#section-6.10.1">Section 6.10.1</a> defines a procedure, the "BitStringLength |
| 1819 |
Compatibility Check", that can be used to detect the possibility of |
| 1820 |
such black holes. |
| 1821 |
|
| 1822 |
However, failure of the BitStringLength Compatibility Check does not |
| 1823 |
necessarily result in the creation of black holes; <a href="#section-6.10.2">Section 6.10.2</a> |
| 1824 |
specifies OPTIONAL procedures that allow BIER forwarding to proceed |
| 1825 |
without black holes, even if the BitStringLength Compatibility Check |
| 1826 |
fails. |
| 1827 |
|
| 1828 |
If the procedures of <a href="#section-6.10.2">Section 6.10.2</a> are not deployed but the |
| 1829 |
BitStringLength Compatibility Check fails at some BFIR, the BFIR has |
| 1830 |
two choices: |
| 1831 |
|
| 1832 |
o Create BIER packets with the provisioned Imposition |
| 1833 |
BitStringLength, even though the packets may not be able to reach |
| 1834 |
all the BFERs identified in their BitStrings. |
| 1835 |
|
| 1836 |
o Use an Imposition BitStringLength that passes the Compatibility |
| 1837 |
Check (assuming that there is one), even if this is not the |
| 1838 |
provisioned Imposition BitStringLength. |
| 1839 |
|
| 1840 |
|
| 1841 |
|
| 1842 |
|
| 1843 |
|
| 1844 |
<span class="grey">Wijnands, et al. Experimental [Page 33]</span></pre> |
| 1845 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-34" ></span> |
| 1846 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1847 |
|
| 1848 |
|
| 1849 |
<a href="#section-6.10.1">Section 6.10.1</a> discusses the implications of making one or the other |
| 1850 |
of these choices. |
| 1851 |
|
| 1852 |
There will be times when an operator wishes to change the |
| 1853 |
BitStringLengths used in a particular BIER domain. <a href="#section-6.10.3">Section 6.10.3</a> |
| 1854 |
specifies a simple procedure that can be used to transition a BIER |
| 1855 |
domain from one BitStringLength to another. |
| 1856 |
|
| 1857 |
<span class="h4"><a class="selflink" id="section-6.10.1" href="#section-6.10.1">6.10.1</a>. BitStringLength Compatibility Check</span> |
| 1858 |
|
| 1859 |
When a BFIR needs to encapsulate a packet, the BFIR first assigns the |
| 1860 |
packet to a sub-domain. The BFIR then chooses an Imposition |
| 1861 |
BitStringLength L for the packet. The choice of Imposition |
| 1862 |
BitStringLength is determined by provisioning. However, the BFIR |
| 1863 |
should also perform the BitStringLength Compatibility Check defined |
| 1864 |
below. |
| 1865 |
|
| 1866 |
The combination of sub-domain S and Imposition BitStringLength L |
| 1867 |
passes the BitStringLength Compatibility Check if and only if the |
| 1868 |
following condition holds: |
| 1869 |
|
| 1870 |
Every BFR that has advertised its membership in sub-domain S has |
| 1871 |
also advertised that it is using Disposition BitStringLength L |
| 1872 |
(and possibly other BitStringLengths as well) in that sub-domain. |
| 1873 |
(If MPLS encapsulation (Section 2.1 of [<a href="#ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>]) is |
| 1874 |
being used, this means that every BFR that is advertising a label |
| 1875 |
for sub-domain S is advertising a label for the combination of |
| 1876 |
sub-domain S and Disposition BitStringLength L.) |
| 1877 |
|
| 1878 |
If a BFIR has been provisioned to use a particular Imposition |
| 1879 |
BitStringLength and a particular sub-domain for some set of packets, |
| 1880 |
and if that combination of Imposition BitStringLength and sub-domain |
| 1881 |
does not pass the BitStringLength Compatibility Check, the BFIR |
| 1882 |
SHOULD log this fact as an error. It then has the following two |
| 1883 |
choices about what to do with the packets: |
| 1884 |
|
| 1885 |
1. The BFIR MAY use the provisioned Imposition BitStringLength |
| 1886 |
anyway. If the procedure of either option 2 or option 3 of |
| 1887 |
<a href="#section-6.10.2">Section 6.10.2</a> is deployed, this will not cause black holes and |
| 1888 |
may actually be the optimal result. It should be understood, |
| 1889 |
though, that the BFIR cannot determine by signaling whether those |
| 1890 |
procedures have been deployed. |
| 1891 |
|
| 1892 |
2. If the BFIR is capable of using an Imposition BitStringLength |
| 1893 |
that does pass the BitStringLength Compatibility Check for the |
| 1894 |
particular sub-domain, the BFIR MAY use that Imposition |
| 1895 |
BitStringLength instead. |
| 1896 |
|
| 1897 |
|
| 1898 |
|
| 1899 |
|
| 1900 |
<span class="grey">Wijnands, et al. Experimental [Page 34]</span></pre> |
| 1901 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-35" ></span> |
| 1902 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1903 |
|
| 1904 |
|
| 1905 |
Which of these two choices to make is itself determined by |
| 1906 |
provisioning. |
| 1907 |
|
| 1908 |
Note that discarding the packets is not one of the allowable choices. |
| 1909 |
Suppose, for example, that all the BFIRs are provisioned to use |
| 1910 |
Imposition BitStringLength L for a particular sub-domain S but one |
| 1911 |
BFR has not been provisioned to use Disposition BitStringLength L for |
| 1912 |
sub-domain S. This will cause the BitStringLength Compatibility |
| 1913 |
Check to fail. If the BFIR sends packets with BitStringLength L and |
| 1914 |
sub-domain S, the mis-provisioned BFR will not be able to forward |
| 1915 |
those packets, and thus the packets may only be able to reach a |
| 1916 |
subset of the BFERs to which they are destined. However, this is |
| 1917 |
still better than having the BFIRs drop the packets; if the BFIRs |
| 1918 |
discard the packets, the packets won't reach any of the BFERs to |
| 1919 |
which they are destined at all. |
| 1920 |
|
| 1921 |
If the procedures of <a href="#section-6.10.2">Section 6.10.2</a> have not been deployed, choice 2 |
| 1922 |
above might seem like a better option. However, there might not be |
| 1923 |
any Imposition BitStringLength that a given BFIR can use that also |
| 1924 |
passes the BitStringLength Compatibility Check. If it is desired to |
| 1925 |
use choice 2 in a particular deployment, then there should be a |
| 1926 |
"Fallback Disposition BitStringLength" (call it "F") such that: |
| 1927 |
|
| 1928 |
o Every BFR advertises that it uses BitStringLength F as a |
| 1929 |
Disposition BitStringLength for every sub-domain, and |
| 1930 |
|
| 1931 |
o If a BFIR is provisioned to use Imposition BitStringLength X and |
| 1932 |
Imposition sub-domain S for a certain class of packets but the |
| 1933 |
BitStringLength Compatibility Check fails for the combination of |
| 1934 |
BitStringLength X and sub-domain S, then the BFIR will fall back |
| 1935 |
to using BitStringLength F as the Imposition BitStringLength |
| 1936 |
whenever the Imposition sub-domain is S. |
| 1937 |
|
| 1938 |
It is RECOMMENDED that the value of F be the default BitStringLength |
| 1939 |
for the encapsulation being used. |
| 1940 |
|
| 1941 |
|
| 1942 |
|
| 1943 |
|
| 1944 |
|
| 1945 |
|
| 1946 |
|
| 1947 |
|
| 1948 |
|
| 1949 |
|
| 1950 |
|
| 1951 |
|
| 1952 |
|
| 1953 |
|
| 1954 |
|
| 1955 |
|
| 1956 |
<span class="grey">Wijnands, et al. Experimental [Page 35]</span></pre> |
| 1957 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-36" ></span> |
| 1958 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 1959 |
|
| 1960 |
|
| 1961 |
<span class="h4"><a class="selflink" id="section-6.10.2" href="#section-6.10.2">6.10.2</a>. Handling BitStringLength Mismatches</span> |
| 1962 |
|
| 1963 |
Suppose that a packet has been BIER-encapsulated with a |
| 1964 |
BitStringLength value of X and that the packet has arrived at BFR-A. |
| 1965 |
Now suppose that according to the routing underlay the next hop is |
| 1966 |
BFR-B, but BFR-B is not using X as one of its Disposition |
| 1967 |
BitStringLengths. What should BFR-A do with the packet? BFR-A has |
| 1968 |
three options. It MUST do one of the three, but the choice of which |
| 1969 |
procedure to follow is a local matter. The three options are: |
| 1970 |
|
| 1971 |
1. BFR-A MAY discard the packet. |
| 1972 |
|
| 1973 |
2. BFR-A MAY re-encapsulate the packet, using a BIER header whose |
| 1974 |
BitStringLength value is supported by BFR-B. |
| 1975 |
|
| 1976 |
Note that if BFR-B only uses Disposition BitStringLength values |
| 1977 |
that are smaller than the BitStringLength value of the packet, |
| 1978 |
this may require creating additional copies of the packet. |
| 1979 |
Whether additional copies actually have to be created depends |
| 1980 |
upon the bits that are actually set in the original packet's |
| 1981 |
BitString. |
| 1982 |
|
| 1983 |
3. BFR-A MAY treat BFR-B as if BFR-B did not support BIER at all, |
| 1984 |
in which case BFR-A applies the rules of <a href="#section-6.9">Section 6.9</a>. |
| 1985 |
|
| 1986 |
Note that there is no signaling that enables a BFR to advertise which |
| 1987 |
of the three options it will use. |
| 1988 |
|
| 1989 |
Option 2 can be useful if there is a region of the BIER domain where |
| 1990 |
the BFRs are capable of using a long BitStringLength as well as a |
| 1991 |
region where the BFRs are only capable of using a shorter |
| 1992 |
BitStringLength. |
| 1993 |
|
| 1994 |
<span class="h4"><a class="selflink" id="section-6.10.3" href="#section-6.10.3">6.10.3</a>. Transitioning from One BitStringLength to Another</span> |
| 1995 |
|
| 1996 |
Suppose one wants to migrate the BitStringLength used in a particular |
| 1997 |
BIER domain from one value (X) to another value (Y). The following |
| 1998 |
migration procedure can be used. This procedure allows the BFRs to |
| 1999 |
be reprovisioned one at a time and does not require a "flag day". |
| 2000 |
|
| 2001 |
1. Upgrade all the BFRs in the domain so that they use both value X |
| 2002 |
and value Y as their Disposition BitStringLengths. |
| 2003 |
|
| 2004 |
2. Reprovision the BFIRs so that they use BitStringLength value Y as |
| 2005 |
the Imposition BitStringLength. |
| 2006 |
|
| 2007 |
3. One may then optionally reprovision all the BFRs so that they no |
| 2008 |
longer use Disposition BitStringLength X. |
| 2009 |
|
| 2010 |
|
| 2011 |
|
| 2012 |
<span class="grey">Wijnands, et al. Experimental [Page 36]</span></pre> |
| 2013 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-37" ></span> |
| 2014 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2015 |
|
| 2016 |
|
| 2017 |
<span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. Operational Considerations</span> |
| 2018 |
|
| 2019 |
BIER offers a radical simplification over current IP multicast |
| 2020 |
operations: no tree-building control plane, no per-flow forwarding |
| 2021 |
state, no Reverse Path Forwarding (RPF), no Rendezvous Point (RP), |
| 2022 |
etc. BIER packet forwarding/replication is along the unicast paths |
| 2023 |
to each bit position set in the packet, ensuring that the |
| 2024 |
encapsulated multicast packets follow the same path as unicast to |
| 2025 |
each set bit in the header. The BIER FIB can be derived from the |
| 2026 |
SPF-calculated unicast FIB or from any other forwarding-path |
| 2027 |
calculation in or out of band. Each bit will follow this unicast |
| 2028 |
path from the entry point of the BIER domain to the edge device with |
| 2029 |
that assigned bit. |
| 2030 |
|
| 2031 |
Due to these differences, operational expectations from traditional |
| 2032 |
multicast solutions do not apply to a BIER domain. There is no |
| 2033 |
granular per-flow state at each node defining a tree. Monitoring |
| 2034 |
flows at the forwarding-plane level ((S,G) entries) is not provided |
| 2035 |
in a BIER node. BIER FIB packet counters may be maintained for |
| 2036 |
BFR-ids or next-hop neighbors. Any flow-based metrics will require |
| 2037 |
deeper packet inspection; this topic is outside the scope of this |
| 2038 |
document. In this way, BIER is again more like unicast. |
| 2039 |
|
| 2040 |
It is this reduction in state that allows for one of the key |
| 2041 |
operational benefits of BIER: deterministic convergence. The BIER |
| 2042 |
FIB can converge immediately after the unicast FIB regardless of how |
| 2043 |
many multicast flows are transiting the links. Careful monitoring of |
| 2044 |
(S,G) utilization is not required within a BIER domain. |
| 2045 |
|
| 2046 |
<span class="h3"><a class="selflink" id="section-7.1" href="#section-7.1">7.1</a>. Configuration</span> |
| 2047 |
|
| 2048 |
A BIER domain requires that each edge node (BFER) be given a unique |
| 2049 |
bit position in the BIER mask (BFR-id). The BFR-id must be |
| 2050 |
configured on each BFER and associated with a unique IP address of |
| 2051 |
that BFER. Any existing manual or automated configuration tools must |
| 2052 |
provide access to BIER-specific configuration. The association of |
| 2053 |
the BFR-id with a unique address of the BFER to which it is assigned |
| 2054 |
must also be advertised into the IGP of the BIER domain. This may be |
| 2055 |
implied from the BIER configuration or require IGP-specific |
| 2056 |
configuration. This document does not dictate any specific |
| 2057 |
configuration methodology. |
| 2058 |
|
| 2059 |
<span class="h2"><a class="selflink" id="section-8" href="#section-8">8</a>. IANA Considerations</span> |
| 2060 |
|
| 2061 |
This document does not require any IANA actions. |
| 2062 |
|
| 2063 |
|
| 2064 |
|
| 2065 |
|
| 2066 |
|
| 2067 |
|
| 2068 |
<span class="grey">Wijnands, et al. Experimental [Page 37]</span></pre> |
| 2069 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-38" ></span> |
| 2070 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2071 |
|
| 2072 |
|
| 2073 |
<span class="h2"><a class="selflink" id="section-9" href="#section-9">9</a>. Security Considerations</span> |
| 2074 |
|
| 2075 |
When BIER is paired with a particular multicast flow overlay, it |
| 2076 |
inherits the security considerations of that layer. Similarly, when |
| 2077 |
BIER is paired with a particular routing underlay, it inherits the |
| 2078 |
security considerations of that layer. |
| 2079 |
|
| 2080 |
If the BIER encapsulation of a particular packet specifies an SI or a |
| 2081 |
BitString other than the one intended by the BFIR, the packet is |
| 2082 |
likely to be misdelivered. If the BIER encapsulation of a packet is |
| 2083 |
modified (through error or malfeasance) in a way other than that |
| 2084 |
specified in this document, the packet may be misdelivered. Some |
| 2085 |
modifications of the BIER encapsulation, e.g., setting every bit in |
| 2086 |
the BitString, may result in (intentional or unintentional) |
| 2087 |
denial-of-service (DoS) attacks. |
| 2088 |
|
| 2089 |
If a BFIR is compromised, it may impose a BIER encapsulation with all |
| 2090 |
the bits in the BitString set; this would also result in a DoS |
| 2091 |
attack. |
| 2092 |
|
| 2093 |
Every BFR MUST be provisioned to know which of its interfaces lead to |
| 2094 |
a BIER domain and which do not. BIER-encapsulated packets MUST NOT |
| 2095 |
be accepted from outside the BIER domain. (Reception of |
| 2096 |
BIER-encapsulated packets from outside the BIER domain would create |
| 2097 |
an attack vector for DoS attacks, as an attacker might set all the |
| 2098 |
bits in the BitString.) |
| 2099 |
|
| 2100 |
If two interfaces lead to different BIER domains, the BFR MUST be |
| 2101 |
provisioned to know that those two interfaces lead to different BIER |
| 2102 |
domains. If the provisioning is not correct, BIER-encapsulated |
| 2103 |
packets from one BIER domain may "leak" into another; this is likely |
| 2104 |
to result in misdelivery of packets. |
| 2105 |
|
| 2106 |
DoS attacks may also result from incorrect provisioning (through |
| 2107 |
error or malfeasance) of the BFRs. |
| 2108 |
|
| 2109 |
If the procedures used for advertising BFR-ids and BFR-prefixes are |
| 2110 |
not secure, an attack on those procedures may result in incorrect |
| 2111 |
delivery of BIER-encapsulated packets. |
| 2112 |
|
| 2113 |
|
| 2114 |
|
| 2115 |
|
| 2116 |
|
| 2117 |
|
| 2118 |
|
| 2119 |
|
| 2120 |
|
| 2121 |
|
| 2122 |
|
| 2123 |
|
| 2124 |
<span class="grey">Wijnands, et al. Experimental [Page 38]</span></pre> |
| 2125 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-39" ></span> |
| 2126 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2127 |
|
| 2128 |
|
| 2129 |
<span class="h2"><a class="selflink" id="section-10" href="#section-10">10</a>. References</span> |
| 2130 |
|
| 2131 |
<span class="h3"><a class="selflink" id="section-10.1" href="#section-10.1">10.1</a>. Normative References</span> |
| 2132 |
|
| 2133 |
[<a id="ref-RFC2119">RFC2119</a>] Bradner, S., "Key words for use in RFCs to Indicate |
| 2134 |
Requirement Levels", <a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a>, <a href="./rfc2119">RFC 2119</a>, |
| 2135 |
DOI 10.17487/RFC2119, March 1997, |
| 2136 |
<<a href="https://www.rfc-editor.org/info/rfc2119">https://www.rfc-editor.org/info/rfc2119</a>>. |
| 2137 |
|
| 2138 |
[<a id="ref-RFC3443">RFC3443</a>] Agarwal, P. and B. Akyol, "Time To Live (TTL) Processing |
| 2139 |
in Multi-Protocol Label Switching (MPLS) Networks", |
| 2140 |
<a href="./rfc3443">RFC 3443</a>, DOI 10.17487/RFC3443, January 2003, |
| 2141 |
<<a href="https://www.rfc-editor.org/info/rfc3443">https://www.rfc-editor.org/info/rfc3443</a>>. |
| 2142 |
|
| 2143 |
[<a id="ref-RFC8174">RFC8174</a>] Leiba, B., "Ambiguity of Uppercase vs Lowercase in |
| 2144 |
<a href="./rfc2119">RFC 2119</a> Key Words", <a href="https://www.rfc-editor.org/bcp/bcp14">BCP 14</a>, <a href="./rfc8174">RFC 8174</a>, |
| 2145 |
DOI 10.17487/RFC8174, May 2017, |
| 2146 |
<<a href="https://www.rfc-editor.org/info/rfc8174">https://www.rfc-editor.org/info/rfc8174</a>>. |
| 2147 |
|
| 2148 |
<span class="h3"><a class="selflink" id="section-10.2" href="#section-10.2">10.2</a>. Informative References</span> |
| 2149 |
|
| 2150 |
[<a id="ref-BGP_BIER_EXTENSIONS">BGP_BIER_EXTENSIONS</a>] |
| 2151 |
Xu, X., Ed., Chen, M., Patel, K., Wijnands, IJ., and A. |
| 2152 |
Przygienda, "BGP Extensions for BIER", Work in Progress, |
| 2153 |
<a href="./draft-ietf-bier-idr-extensions-03">draft-ietf-bier-idr-extensions-03</a>, August 2017. |
| 2154 |
|
| 2155 |
[<a id="ref-BIER_MVPN">BIER_MVPN</a>] |
| 2156 |
Rosen, E., Ed., Sivakumar, M., Aldrin, S., Dolganow, A., |
| 2157 |
and T. Przygienda, "Multicast VPN Using BIER", Work in |
| 2158 |
Progress, <a href="./draft-ietf-bier-mvpn-09">draft-ietf-bier-mvpn-09</a>, November 2017. |
| 2159 |
|
| 2160 |
[<a id="ref-Boivie_Feldman">Boivie_Feldman</a>] |
| 2161 |
Boivie, R. and N. Feldman, <a style="text-decoration: none" href='https://www.google.com/search?sitesearch=datatracker.ietf.org%2Fdoc%2Fhtml%2F&q=inurl:draft-+%22Small+Group+Multicast%22'>"Small Group Multicast"</a>, Work |
| 2162 |
in Progress, <a href="./draft-boivie-sgm-02">draft-boivie-sgm-02</a>, February 2001. |
| 2163 |
|
| 2164 |
[<a id="ref-ISIS_BIER_EXTENSIONS">ISIS_BIER_EXTENSIONS</a>] |
| 2165 |
Ginsberg, L., Ed., Przygienda, A., Aldrin, S., and J. |
| 2166 |
Zhang, "BIER Support via ISIS", Work in Progress, |
| 2167 |
<a href="./draft-ietf-bier-isis-extensions-06">draft-ietf-bier-isis-extensions-06</a>, October 2017. |
| 2168 |
|
| 2169 |
[<a id="ref-MPLS_BIER_ENCAPS">MPLS_BIER_ENCAPS</a>] |
| 2170 |
Wijnands, IJ., Ed., Rosen, E., Ed., Dolganow, A., |
| 2171 |
Tantsura, J., Aldrin, S., and I. Meilik, "Encapsulation |
| 2172 |
for Bit Index Explicit Replication in MPLS and non-MPLS |
| 2173 |
Networks", Work in Progress, <a href="./draft-ietf-bier-mpls-encapsulation-12">draft-ietf-bier-mpls-</a> |
| 2174 |
<a href="./draft-ietf-bier-mpls-encapsulation-12">encapsulation-12</a>, October 2017. |
| 2175 |
|
| 2176 |
|
| 2177 |
|
| 2178 |
|
| 2179 |
|
| 2180 |
<span class="grey">Wijnands, et al. Experimental [Page 39]</span></pre> |
| 2181 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-40" ></span> |
| 2182 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2183 |
|
| 2184 |
|
| 2185 |
[<a id="ref-OSPF_BIER_EXTENSIONS">OSPF_BIER_EXTENSIONS</a>] |
| 2186 |
Psenak, P., Ed., Kumar, N., Wijnands, IJ., Dolganow, A., |
| 2187 |
Przygienda, T., Zhang, J., and S. Aldrin, "OSPF Extensions |
| 2188 |
for BIER", Work in Progress, <a href="./draft-ietf-bier-ospf-bier-extensions-09">draft-ietf-bier-ospf-bier-</a> |
| 2189 |
<a href="./draft-ietf-bier-ospf-bier-extensions-09">extensions-09</a>, October 2017. |
| 2190 |
|
| 2191 |
[<a id="ref-RFC6513">RFC6513</a>] Rosen, E., Ed., and R. Aggarwal, Ed., "Multicast in |
| 2192 |
MPLS/BGP IP VPNs", <a href="./rfc6513">RFC 6513</a>, DOI 10.17487/RFC6513, |
| 2193 |
February 2012, <<a href="https://www.rfc-editor.org/info/rfc6513">https://www.rfc-editor.org/info/rfc6513</a>>. |
| 2194 |
|
| 2195 |
[<a id="ref-RFC6514">RFC6514</a>] Aggarwal, R., Rosen, E., Morin, T., and Y. Rekhter, "BGP |
| 2196 |
Encodings and Procedures for Multicast in MPLS/BGP IP |
| 2197 |
VPNs", <a href="./rfc6514">RFC 6514</a>, DOI 10.17487/RFC6514, February 2012, |
| 2198 |
<<a href="https://www.rfc-editor.org/info/rfc6514">https://www.rfc-editor.org/info/rfc6514</a>>. |
| 2199 |
|
| 2200 |
Acknowledgements |
| 2201 |
|
| 2202 |
The authors wish to thank Rajiv Asati, Alia Atlas, John Bettink, Ross |
| 2203 |
Callon (who contributed much of the text on deterministic ECMP), |
| 2204 |
Nagendra Kumar, Christian Martin, Neale Ranns, Albert Tian, Ramji |
| 2205 |
Vaithianathan, Xiaohu Xu, and Jeffrey Zhang for their ideas and |
| 2206 |
contributions to this work. |
| 2207 |
|
| 2208 |
The authors also wish to thank Sue Hares, Victor Kuarsingh, and Dan |
| 2209 |
Romascanu for their reviews of this document. |
| 2210 |
|
| 2211 |
|
| 2212 |
|
| 2213 |
|
| 2214 |
|
| 2215 |
|
| 2216 |
|
| 2217 |
|
| 2218 |
|
| 2219 |
|
| 2220 |
|
| 2221 |
|
| 2222 |
|
| 2223 |
|
| 2224 |
|
| 2225 |
|
| 2226 |
|
| 2227 |
|
| 2228 |
|
| 2229 |
|
| 2230 |
|
| 2231 |
|
| 2232 |
|
| 2233 |
|
| 2234 |
|
| 2235 |
|
| 2236 |
<span class="grey">Wijnands, et al. Experimental [Page 40]</span></pre> |
| 2237 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-41" ></span> |
| 2238 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2239 |
|
| 2240 |
|
| 2241 |
Contributors |
| 2242 |
|
| 2243 |
The following people contributed significantly to the content of this |
| 2244 |
document and should be considered co-authors: |
| 2245 |
|
| 2246 |
Gregory Cauchie |
| 2247 |
Bouygues Telecom |
| 2248 |
Email: [email protected] |
| 2249 |
|
| 2250 |
Mach(Guoyi) Chen |
| 2251 |
Huawei |
| 2252 |
Email: [email protected] |
| 2253 |
|
| 2254 |
Arkadiy Gulko |
| 2255 |
Thomson Reuters |
| 2256 |
195 Broadway |
| 2257 |
New York, NY 10007 |
| 2258 |
United States of America |
| 2259 |
Email: [email protected] |
| 2260 |
|
| 2261 |
Wim Henderickx |
| 2262 |
Nokia |
| 2263 |
Copernicuslaan 50 |
| 2264 |
Antwerp 2018 |
| 2265 |
Belgium |
| 2266 |
Email: [email protected] |
| 2267 |
|
| 2268 |
Martin Horneffer |
| 2269 |
Deutsche Telekom |
| 2270 |
Hammer Str. 216-226 |
| 2271 |
Muenster 48153 |
| 2272 |
Germany |
| 2273 |
Email: [email protected] |
| 2274 |
|
| 2275 |
Luay Jalil |
| 2276 |
Verizon |
| 2277 |
1201 East Arapaho Rd. |
| 2278 |
Richardson, TX 75081 |
| 2279 |
United States of America |
| 2280 |
Email: [email protected] |
| 2281 |
|
| 2282 |
Uwe Joorde |
| 2283 |
Deutsche Telekom |
| 2284 |
Hammer Str. 216-226 |
| 2285 |
Muenster D-48153 |
| 2286 |
Germany |
| 2287 |
Email: [email protected] |
| 2288 |
|
| 2289 |
|
| 2290 |
|
| 2291 |
|
| 2292 |
<span class="grey">Wijnands, et al. Experimental [Page 41]</span></pre> |
| 2293 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-42" ></span> |
| 2294 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2295 |
|
| 2296 |
|
| 2297 |
Greg Shepherd |
| 2298 |
Cisco Systems |
| 2299 |
170 West Tasman Drive |
| 2300 |
San Jose, CA 95134 |
| 2301 |
United States of America |
| 2302 |
Email: [email protected] |
| 2303 |
|
| 2304 |
Jeff Tantsura |
| 2305 |
Email: [email protected] |
| 2306 |
|
| 2307 |
|
| 2308 |
|
| 2309 |
|
| 2310 |
|
| 2311 |
|
| 2312 |
|
| 2313 |
|
| 2314 |
|
| 2315 |
|
| 2316 |
|
| 2317 |
|
| 2318 |
|
| 2319 |
|
| 2320 |
|
| 2321 |
|
| 2322 |
|
| 2323 |
|
| 2324 |
|
| 2325 |
|
| 2326 |
|
| 2327 |
|
| 2328 |
|
| 2329 |
|
| 2330 |
|
| 2331 |
|
| 2332 |
|
| 2333 |
|
| 2334 |
|
| 2335 |
|
| 2336 |
|
| 2337 |
|
| 2338 |
|
| 2339 |
|
| 2340 |
|
| 2341 |
|
| 2342 |
|
| 2343 |
|
| 2344 |
|
| 2345 |
|
| 2346 |
|
| 2347 |
|
| 2348 |
<span class="grey">Wijnands, et al. Experimental [Page 42]</span></pre> |
| 2349 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-43" ></span> |
| 2350 |
<span class="grey"><a href="./rfc8279">RFC 8279</a> Multicast with BIER November 2017</span> |
| 2351 |
|
| 2352 |
|
| 2353 |
Authors' Addresses |
| 2354 |
|
| 2355 |
IJsbrand Wijnands (editor) |
| 2356 |
Cisco Systems, Inc. |
| 2357 |
De Kleetlaan 6a |
| 2358 |
Diegem 1831 |
| 2359 |
Belgium |
| 2360 |
|
| 2361 |
Email: [email protected] |
| 2362 |
|
| 2363 |
|
| 2364 |
Eric C. Rosen (editor) |
| 2365 |
Juniper Networks, Inc. |
| 2366 |
10 Technology Park Drive |
| 2367 |
Westford, Massachusetts 01886 |
| 2368 |
United States of America |
| 2369 |
|
| 2370 |
Email: [email protected] |
| 2371 |
|
| 2372 |
|
| 2373 |
Andrew Dolganow |
| 2374 |
Nokia |
| 2375 |
438B Alexandra Rd #08-07/10 |
| 2376 |
Alexandra Technopark |
| 2377 |
Singapore 119968 |
| 2378 |
Singapore |
| 2379 |
|
| 2380 |
Email: [email protected] |
| 2381 |
|
| 2382 |
|
| 2383 |
Tony Przygienda |
| 2384 |
Juniper Networks, Inc. |
| 2385 |
1194 N. Mathilda Ave. |
| 2386 |
Sunnyvale, California 94089 |
| 2387 |
United States of America |
| 2388 |
|
| 2389 |
Email: [email protected] |
| 2390 |
|
| 2391 |
|
| 2392 |
Sam K. Aldrin |
| 2393 |
Google, Inc. |
| 2394 |
1600 Amphitheatre Parkway |
| 2395 |
Mountain View, California 94043 |
| 2396 |
United States of America |
| 2397 |
|
| 2398 |
Email: [email protected] |
| 2399 |
|
| 2400 |
|
| 2401 |
|
| 2402 |
|
| 2403 |
|
| 2404 |
Wijnands, et al. Experimental [Page 43] |
| 2405 |
</pre> |