| 1 |
<pre>Network Working Group D. Provan |
| 2 |
Request for Comments: 1201 Novell, Inc. |
| 3 |
Obsoletes: RFC <a href="./rfc1051">1051</a> February 1991 |
| 4 |
|
| 5 |
|
| 6 |
<span class="h1">Transmitting IP Traffic over ARCNET Networks</span> |
| 7 |
|
| 8 |
Status of this Memo |
| 9 |
|
| 10 |
This memo defines a protocol for the transmission of IP and ARP |
| 11 |
packets over the ARCnet Local Area Network. This RFC specifies an |
| 12 |
IAB standards track protocol for the Internet community, and requests |
| 13 |
discussion and suggestions for improvements. Please refer to the |
| 14 |
current edition of the "IAB Official Protocol Standards" for the |
| 15 |
standardization state and status of this protocol. Distribution of |
| 16 |
this memo is unlimited. |
| 17 |
|
| 18 |
<span class="h2"><a class="selflink" id="section-1" href="#section-1">1</a>. Introduction</span> |
| 19 |
|
| 20 |
This memo specifies a method of encapsulating Internet Protocol (IP) |
| 21 |
[<a href="#ref-1" title=""Internet Protocol"">1</a>] and Address Resolution Protocol (ARP) [<a href="#ref-2" title=""An Ethernet Address Resolution Protocol"">2</a>] datagrams for |
| 22 |
transmission across ARCNET [<a href="#ref-3" title=""ARCNET Designer's Handbook"">3</a>] using the "ARCNET Packet Header |
| 23 |
Definition Standard" [<a href="#ref-4" title=""ARCNET Packet Header Definition Standard"">4</a>]. This memo offers a replacement for <a href="./rfc1051">RFC</a> |
| 24 |
<a href="./rfc1051">1051</a>. <a href="./rfc1051">RFC 1051</a> uses an ARCNET framing protocol which limits |
| 25 |
unfragmented IP packets to 508 octets [<a href="#ref-5" title=""A Standard for the Transmission of IP Datagrams and ARP Packets over ARCNET Networks"">5</a>]. |
| 26 |
|
| 27 |
<span class="h2"><a class="selflink" id="section-2" href="#section-2">2</a>. ARCNET Packet Format</span> |
| 28 |
|
| 29 |
In 1989, Apple Computers, Novell, ACTINET Systems, Standard |
| 30 |
Microsystems, and Pure Data Research agreed to use the ARCNET |
| 31 |
datalink protocol defined in "ARCNET Packet Header Definition |
| 32 |
Standard" [<a href="#ref-4" title=""ARCNET Packet Header Definition Standard"">4</a>]. We'll begin with a brief description of that |
| 33 |
protocol. |
| 34 |
|
| 35 |
<span class="h3"><a class="selflink" id="section-2.1" href="#section-2.1">2.1</a>. ARCNET Framing</span> |
| 36 |
|
| 37 |
ARCNET hardware supports two types of frames: short frames, which are |
| 38 |
always 256 octets long, and long frames, which are always 512 octets |
| 39 |
long. All frames begin with a hardware header and end with the |
| 40 |
client's data preceded by a software header. Software places padding |
| 41 |
in the middle of the packet between the hardware header and the |
| 42 |
software header to make the frame the appropriate fixed length. |
| 43 |
Unbeknown to the software, the hardware removes this padding during |
| 44 |
transmission. |
| 45 |
|
| 46 |
Short frames can hold from 0 to 249 octets of client data. Long |
| 47 |
frames can hold from 253 to 504 octets of client data. To handle |
| 48 |
frames with 250, 251, or 252 octets of data, the datalink protocol |
| 49 |
|
| 50 |
|
| 51 |
|
| 52 |
<span class="grey">Provan [Page 1]</span></pre> |
| 53 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-2" ></span> |
| 54 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 55 |
|
| 56 |
|
| 57 |
introduces a third frame type: the exception frame. |
| 58 |
|
| 59 |
These three frame formats are shown here. Except as noted, each |
| 60 |
block represents one octet. |
| 61 |
|
| 62 |
|
| 63 |
Short Frame Long Frame Exception Frame |
| 64 |
|
| 65 |
+---------------+ +---------------+ +---------------+ |
| 66 |
| source | | source | | source | |
| 67 |
+---------------+ +---------------+ +---------------+ |
| 68 |
| destination | | destination | | destination | |
| 69 |
+---------------+ +---------------+ +---------------+ |
| 70 |
| offset | | 0 | | 0 | |
| 71 |
+---------------+ +---------------+ +---------------+ |
| 72 |
. unused . | offset | | offset | |
| 73 |
. (offset - 3 . +---------------+ +---------------+ |
| 74 |
. octets) . . unused . . unused . |
| 75 |
+---------------+ . (offset - 4 . . (offset - 4 . |
| 76 |
| protocol ID | . octets) . . octets) . |
| 77 |
+---------------+ +---------------+ +---------------+ |
| 78 |
| split flag | | protocol ID | | protocol ID | |
| 79 |
+---------------+ +---------------+ +---------------+ |
| 80 |
| sequence | | split flag | | flag: FF hex | |
| 81 |
+ number + +---------------+ +---------------+ |
| 82 |
| (2 octets) | | sequence | | padding: 0xFF | |
| 83 |
+---------------+ + number + +---------------+ |
| 84 |
. . | (2 octets) | | padding: 0xFF | |
| 85 |
. client data . +---------------+ +---------------+ |
| 86 |
. (256 - offset . . . | (protocol ID) | |
| 87 |
. - 4 octets) . . . +---------------+ |
| 88 |
. . . . | split flag | |
| 89 |
+---------------+ . . +---------------+ |
| 90 |
. . | sequence | |
| 91 |
. client data . + number + |
| 92 |
. (512 - offset . | (2 octets) | |
| 93 |
. - 4 octets) . +---------------+ |
| 94 |
. . . . |
| 95 |
. . . client data . |
| 96 |
. . . (512 - offset . |
| 97 |
. . . - 8 octets) . |
| 98 |
. . . . |
| 99 |
+---------------+ +---------------+ |
| 100 |
|
| 101 |
These packet formats are presented as software would see them |
| 102 |
through ARCNET hardware. [<a href="#ref-3" title=""ARCNET Designer's Handbook"">3</a>] refers to this as the "buffer |
| 103 |
format". The actual format of packets on the wire is a little |
| 104 |
different: the destination ID is duplicated, the padding between |
| 105 |
|
| 106 |
|
| 107 |
|
| 108 |
<span class="grey">Provan [Page 2]</span></pre> |
| 109 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-3" ></span> |
| 110 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 111 |
|
| 112 |
|
| 113 |
the offset field and the protocol ID field is not transmitted, and |
| 114 |
there's some hardware framing information. In addition, the |
| 115 |
hardware transmits special packets for buffer allocation and |
| 116 |
reception acknowledgement which are not described here [<a href="#ref-3" title=""ARCNET Designer's Handbook"">3</a>]. |
| 117 |
|
| 118 |
<span class="h3"><a class="selflink" id="section-2.2" href="#section-2.2">2.2</a>. Datalink Layer Fragmentation</span> |
| 119 |
|
| 120 |
ARCNET hardware limits individual frames to 512 octets, which allows |
| 121 |
504 octets of client data. This ARCNET datalink protocol allows the |
| 122 |
datalink layer to break packets into as many as 120 fragments for |
| 123 |
transmission. This allows ARCNET clients to transmit up to 60,480 |
| 124 |
octets in each packet. |
| 125 |
|
| 126 |
The "split flag" describes datalink layer packet fragments. There |
| 127 |
are three cases: an unfragmented packet, the first fragment of a |
| 128 |
fragmented packet, and any other fragment of a fragmented packet. |
| 129 |
|
| 130 |
Unfragmented packets always have a split flag of zero. |
| 131 |
|
| 132 |
The first fragment of a fragmented packet has a split flag equal to |
| 133 |
((T-2)*2)+1, where T is the total number of fragments to expect for |
| 134 |
the packet. |
| 135 |
|
| 136 |
Subsequent fragments of a fragmented packet have a split flag equal |
| 137 |
to ((N-1)*2), where N is the number of this fragment. For example, |
| 138 |
the fourth fragment of a packet will always have the split flag value |
| 139 |
of six ( (4-1)*2 ). |
| 140 |
|
| 141 |
The receiving station can identify the last fragment of a packet |
| 142 |
because the value of its 8-bit split flag will be one greater than |
| 143 |
the split flag of the first fragment of the packet. |
| 144 |
|
| 145 |
A previous version of this ARCNET datalink protocol definition |
| 146 |
only allowed packets which could be contained in two fragments. |
| 147 |
In this older standard, the only legal split flags were 0, 1, and |
| 148 |
2. Compatibility with this older standard can be maintained by |
| 149 |
configuring the maximum client data length to 1008 octets. |
| 150 |
|
| 151 |
No more that 120 fragments are allowed. The highest legal split flag |
| 152 |
value is EE hex. (Notice that the split flag value FF hex is used to |
| 153 |
flag exception packets in what would otherwise be a long packet's |
| 154 |
split flag field.) |
| 155 |
|
| 156 |
All fragments of a single packet carry the same sequence number. |
| 157 |
|
| 158 |
<span class="h3"><a class="selflink" id="section-2.3" href="#section-2.3">2.3</a>. Datalink Layer Reassembly</span> |
| 159 |
|
| 160 |
The previous section provides enough information to implement |
| 161 |
|
| 162 |
|
| 163 |
|
| 164 |
<span class="grey">Provan [Page 3]</span></pre> |
| 165 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-4" ></span> |
| 166 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 167 |
|
| 168 |
|
| 169 |
datalink reassembly. To avoid buffer allocation problems during |
| 170 |
reassembly, we recommend allocating enough space for the entire |
| 171 |
reassembled packet when the first fragment arrives. |
| 172 |
|
| 173 |
Since fragments are sent in order, the reassembly procedure can give |
| 174 |
up on a packet if it receives a fragment out of order. There is one |
| 175 |
exception, however. It is possible for successfully received |
| 176 |
fragments to be retransmitted. Reassembly software should ignore |
| 177 |
repetitious fragments without giving up on the packet. |
| 178 |
|
| 179 |
Since fragments will be sent briskly, the reassembly procedure can |
| 180 |
give up on a partially reassembled packet if no additional fragments |
| 181 |
for it arrive within a few seconds. |
| 182 |
|
| 183 |
<span class="h3"><a class="selflink" id="section-2.4" href="#section-2.4">2.4</a>. Datalink Layer Retransmission</span> |
| 184 |
|
| 185 |
For each unicast ARCNET packet, the hardware indicates to the sender |
| 186 |
whether or not the receiver acknowledged the packet. To improve |
| 187 |
reliability, datalink implementations are encouraged to retransmit |
| 188 |
unacknowledged packets or packet fragments. Several retransmissions |
| 189 |
may be necessary. Broadcast packets, however, are never acknowledged |
| 190 |
and, therefore, they should never be retransmitted. |
| 191 |
|
| 192 |
Packets which are successfully received may not be successfully |
| 193 |
acknowledged. Consequently, retransmission by the datalink |
| 194 |
implementation can cause duplicate packets or duplicate fragments. |
| 195 |
Duplicate packets are not a problem for IP or ARP. As mentioned in |
| 196 |
the previous section, ARCNET reassembly support should ignore any |
| 197 |
redundant fragments. |
| 198 |
|
| 199 |
<span class="h2"><a class="selflink" id="section-3" href="#section-3">3</a>. Transmitting IP and ARP Datagrams</span> |
| 200 |
|
| 201 |
IP and ARP datagrams are carried in the client data area of ARCNET |
| 202 |
packets. Datalink support places each datagram in an appropriate |
| 203 |
size ARCNET frame, fragmenting IP datagrams larger than 504 octets |
| 204 |
into multiple frames as described in the previous section. |
| 205 |
|
| 206 |
<span class="h2"><a class="selflink" id="section-4" href="#section-4">4</a>. IP Address Mappings</span> |
| 207 |
|
| 208 |
This section explains how each of the three basic 32-bit internet |
| 209 |
address types are mapped to 8-bit ARCNET addresses. |
| 210 |
|
| 211 |
<span class="h3"><a class="selflink" id="section-4.1" href="#section-4.1">4.1</a>. Unicast Addresses</span> |
| 212 |
|
| 213 |
A unicast IP address is mapped to an 8-bit ARCNET address using ARP |
| 214 |
as specified in [<a href="#ref-2" title=""An Ethernet Address Resolution Protocol"">2</a>]. A later section covers the specific values |
| 215 |
which should be used in ARP packets sent on ARCNET networks. |
| 216 |
|
| 217 |
|
| 218 |
|
| 219 |
|
| 220 |
<span class="grey">Provan [Page 4]</span></pre> |
| 221 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-5" ></span> |
| 222 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 223 |
|
| 224 |
|
| 225 |
It is possible to assign IP addresses such that the last eight |
| 226 |
bits are the same as the 8-bit ARCNET address. This would allow |
| 227 |
direct mapping of IP address to ARCNET address without using a |
| 228 |
discovery protocol. Some implementations might provide this as an |
| 229 |
option, but it is not recommended practice. Although such hard- |
| 230 |
wired mapping is initially appealing, experience shows that ARP is |
| 231 |
a much more flexible and convenient approach which has a very |
| 232 |
small cost. |
| 233 |
|
| 234 |
<span class="h3"><a class="selflink" id="section-4.2" href="#section-4.2">4.2</a>. Broadcast Addresses</span> |
| 235 |
|
| 236 |
All IP broadcast addresses must be mapped to the ARCNET broadcast |
| 237 |
address of 0. |
| 238 |
|
| 239 |
Unlike unicast packets, ARCNET does not attempt to insure delivery |
| 240 |
of broadcast packets, so they may be lost. This will not have a |
| 241 |
major impact on IP since neither IP nor ARP expect all packets to |
| 242 |
be delivered. |
| 243 |
|
| 244 |
<span class="h3"><a class="selflink" id="section-4.3" href="#section-4.3">4.3</a>. Multicast Addresses</span> |
| 245 |
|
| 246 |
Since ARCNET provides no support for multicasts, all IP multicast |
| 247 |
addresses must be mapped to the ARCNET broadcast address of 0. |
| 248 |
|
| 249 |
<span class="h2"><a class="selflink" id="section-5" href="#section-5">5</a>. ARP</span> |
| 250 |
|
| 251 |
The hardware address length is 1 octet for ARP packets sent over |
| 252 |
ARCNET networks. The ARP hardware type for ARCNET is 7. ARP request |
| 253 |
packets are broadcast by directing them to ARCNET broadcast address, |
| 254 |
which is 0. |
| 255 |
|
| 256 |
<span class="h2"><a class="selflink" id="section-6" href="#section-6">6</a>. RARP</span> |
| 257 |
|
| 258 |
Reverse Address Resolution Protocol [<a href="#ref-6" title=""A Reverse Address Resolution Protocol"">6</a>] packets can also be |
| 259 |
transmitted over ARCNET. For the purposes of datalink transmission |
| 260 |
and reception, RARP is identical to ARP and can be handled the same |
| 261 |
way. There are a few differences to notice, however, between RARP |
| 262 |
when running over ARCNET, which has a one octet hardware address, and |
| 263 |
Ethernet, which has a six octet hardware address. |
| 264 |
|
| 265 |
First, there are only 255 different hardware addresses for any given |
| 266 |
ARCNET while there's an very large number of possible Ethernet |
| 267 |
addresses. Second, ARCNET hardware addresses are more likely to be |
| 268 |
duplicated on different ARCNET networks; Ethernet hardware addresses |
| 269 |
will normally be globally unique. Third, an ARCNET hardware address |
| 270 |
is not as constant as an Ethernet address: ARCNET hardware addresses |
| 271 |
are set by switches, not fixed in ROM as they are on Ethernet. |
| 272 |
|
| 273 |
|
| 274 |
|
| 275 |
|
| 276 |
<span class="grey">Provan [Page 5]</span></pre> |
| 277 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-6" ></span> |
| 278 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 279 |
|
| 280 |
|
| 281 |
<span class="h2"><a class="selflink" id="section-7" href="#section-7">7</a>. Maximum Transmission Unit</span> |
| 282 |
|
| 283 |
The maximum IP packet length possible using this encapsulation method |
| 284 |
is 60,480 octets. Since this length is impractical, all ARCNET |
| 285 |
implementations on a given ARCNET network will need to agree on a |
| 286 |
smaller value. Therefore, the maximum packet size MUST be |
| 287 |
configurable in implementations of this specification. |
| 288 |
|
| 289 |
In any case, implementations must be able to send and receive IP |
| 290 |
datagrams up to 576 octets in length, and are strongly encouraged to |
| 291 |
handle IP datagrams up to 1500 octets in length. |
| 292 |
|
| 293 |
Implementations may accept arriving IP datagrams which are larger |
| 294 |
than their configured maximum transmission unit. They are not |
| 295 |
required to discard such datagrams. |
| 296 |
|
| 297 |
To minimize the amount of ARCNET fragmentation, implementations may |
| 298 |
want to aim at an optimum IP packet size of 504 bytes. This avoids |
| 299 |
the overhead of datalink fragmentation, but at the expense of |
| 300 |
increasing the number of IP packets which must be handled by each |
| 301 |
node in the path. In addition to encouraging local applications to |
| 302 |
generate smaller packets, an implementation might also use the TCP |
| 303 |
maximum segment size option to indicate a desire for 464 octet TCP |
| 304 |
segments [<a href="#ref-7" title=""Transmission Control Protocol"">7</a>], or it might announce an IP MTU of 504 octets through |
| 305 |
an MTU discovery mechanism such as [<a href="#ref-8" title=""IP MTU Discovery Options"">8</a>]. These would inform non- |
| 306 |
ARCNET nodes of the smaller optimum packet size. |
| 307 |
|
| 308 |
<span class="h2"><a class="selflink" id="section-8" href="#section-8">8</a>. Assigned Numbers</span> |
| 309 |
|
| 310 |
Datapoint Corporation assigns ARCNET protocol IDs to identify |
| 311 |
different protocols running on the same ARCNET medium. For |
| 312 |
implementations of this specification, Datapoint has assigned 212 |
| 313 |
decimal to IP, 213 decimal to ARP, and 214 decimal to RARP. These |
| 314 |
are not the numbers assigned to the IP encapsulation defined by <a href="./rfc1051">RFC</a> |
| 315 |
<a href="./rfc1051">1051</a> [<a href="#ref-5" title=""A Standard for the Transmission of IP Datagrams and ARP Packets over ARCNET Networks"">5</a>]. Implementations of <a href="./rfc1051">RFC 1051</a> can exist on the same ARCNET |
| 316 |
as implementations of this specification, although the two would not |
| 317 |
be able to communicate with each other. |
| 318 |
|
| 319 |
The Internet Assigned Numbers Authority (IANA) assigns ARP hardware |
| 320 |
type values. It has assigned ARCNET the ARP hardware type of 7 [<a href="#ref-9" title=""Assigned Numbers"">9</a>]. |
| 321 |
|
| 322 |
Acknowledgements |
| 323 |
|
| 324 |
Several people have reviewed this specification and provided useful |
| 325 |
input. I'd like to thank Wesley Hardell at Datapoint and Troy Thomas |
| 326 |
at Novell's Provo office for helping me figure out ARCNET. In |
| 327 |
addition, I particularly appreciate the effort by James VanBokkelen |
| 328 |
at FTP Software who picked on me until all the fuzzy edges were |
| 329 |
|
| 330 |
|
| 331 |
|
| 332 |
<span class="grey">Provan [Page 6]</span></pre> |
| 333 |
<hr class='noprint'/><!--NewPage--><pre class='newpage'><span id="page-7" ></span> |
| 334 |
<span class="grey"><a href="./rfc1201">RFC 1201</a> IP on ARCNET February 1991</span> |
| 335 |
|
| 336 |
|
| 337 |
smoothed out. |
| 338 |
|
| 339 |
The pioneering work in transmitting IP traffic on ARCNET networks was |
| 340 |
done by Philippe Prindeville. |
| 341 |
|
| 342 |
References |
| 343 |
|
| 344 |
[<a id="ref-1">1</a>] Postel, J., "Internet Protocol", <a href="./rfc791">RFC 791</a>, DARPA, September 1981. |
| 345 |
|
| 346 |
[<a id="ref-2">2</a>] Plummer, D., "An Ethernet Address Resolution Protocol", <a href="./rfc826">RFC 826</a>, |
| 347 |
MIT, November 1982. |
| 348 |
|
| 349 |
[<a id="ref-3">3</a>] Datapoint, Corp., "ARCNET Designer's Handbook", Document Number |
| 350 |
61610, 2nd Edition, Datapoint Corporation, 1988. |
| 351 |
|
| 352 |
[<a id="ref-4">4</a>] Novell, Inc., "ARCNET Packet Header Definition Standard", Novell, |
| 353 |
Inc., November 1989. |
| 354 |
|
| 355 |
[<a id="ref-5">5</a>] Prindeville, P., "A Standard for the Transmission of IP Datagrams |
| 356 |
and ARP Packets over ARCNET Networks", <a href="./rfc1051">RFC 1051</a>, McGill |
| 357 |
University, March 1988. |
| 358 |
|
| 359 |
[<a id="ref-6">6</a>] Finlayson, R., Mann, T., Mogul, J., and M. Theimer, "A Reverse |
| 360 |
Address Resolution Protocol", <a href="./rfc903">RFC 903</a>, Stanford, June 1984. |
| 361 |
|
| 362 |
[<a id="ref-7">7</a>] Postel, J., "Transmission Control Protocol", <a href="./rfc793">RFC 793</a>, DARPA, |
| 363 |
September 1981. |
| 364 |
|
| 365 |
[<a id="ref-8">8</a>] Mogul, J., Kent, C., Partridge, C., and K. McCloghrie, "IP MTU |
| 366 |
Discovery Options", <a href="./rfc1063">RFC 1063</a>, DEC, BBN, TWG, July 1988. |
| 367 |
|
| 368 |
[<a id="ref-9">9</a>] Reynolds, J., and J. Postel, "Assigned Numbers", <a href="./rfc1060">RFC 1060</a>, |
| 369 |
USC/Information Sciences Institute, March 1990. |
| 370 |
|
| 371 |
Security Considerations |
| 372 |
|
| 373 |
Security issues are not discussed in this memo. |
| 374 |
|
| 375 |
Author's Address |
| 376 |
|
| 377 |
Don Provan |
| 378 |
Novell, Inc. |
| 379 |
2180 Fortune Drive |
| 380 |
San Jose, California, 95131 |
| 381 |
|
| 382 |
Phone: (408) 473-8440 |
| 383 |
EMail: [email protected] |
| 384 |
|
| 385 |
|
| 386 |
|
| 387 |
|
| 388 |
Provan [Page 7] |
| 389 |
</pre> |