Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!sdd.hp.com!decwrl!sgi!rpw3@rigden.wpd.sgi.com From: rpw3@rigden.wpd.sgi.com (Rob Warnock) Newsgroups: comp.dcom.lans Subject: Re: 802.2 issues Message-ID: <91117@sgi.sgi.com> Date: 14 Mar 91 10:23:40 GMT References: <7160@emory.mathcs.emory.edu> Sender: news@sgi.sgi.com Reply-To: rpw3@sgi.com (Rob Warnock) Organization: Silicon Graphics, Inc., Mountain View, CA Lines: 73 In article <7160@emory.mathcs.emory.edu> km@mathcs.emory.edu writes: +--------------- | 1) The 1985 IEEE 802.2 standard book says that "A number of LSAP code | points have been identified for particular uses. A list of these code | points can be obtained from the IEEE standards office". Does anyone | have a list of these handy? +--------------- See RFC 1060 "Assigned Numbers" for a few of them: IEEE Internet binary binary decimal 00000000 00000000 0 Null LSAP [IEEE] 01000000 00000010 2 Indiv LLC Sublayer Mgt [IEEE] 11000000 00000011 3 Group LLC Sublayer Mgt [IEEE] 00100000 00000100 4 SNA Path Control [IEEE] 01100000 00000110 6 Reserved (DOD IP) [104,JBP] 01110000 00001110 14 PROWAY-LAN [IEEE] 01110010 01001110 78 EIA-RS 511 [IEEE] 01111010 01011110 94 ISI IP [JBP] 01110001 10001110 142 PROWAY-LAN [IEEE] 01010101 10101010 170 SNAP [IEEE] 01111111 11111110 254 ISO DIS 8473 [64,JXJ] 11111111 11111111 255 Global DSAP [IEEE] +--------------- | 2) I know informally about the SAP "AA", which goes with a SNAP | discriminator. Is "AA" one of these code points? Where can I find a | formal description of the "AA" SAP and SNAP? +--------------- Here's an informal one... When 802.2 was being defined, there were some various interests who lobbied succesfully for an "escape hatch" to LLC, which became SNAP. [I've heard various explanations for the acronym, but all of them seem to agree that it's some variation of {Sub-,Sudden,Special} Network Access Protocol.] If the SSAP/DSAP are each 0xAA, then the next three bytes after the LLC header are an OUI (Organization-Unique Identifier) -- I could be wrong, but I think the upper 24 bits of your registered 48-bit MAC address group will do nicely here -- and the two bytes after that are "reserved for organization definition". That is, each "organization" (OUI number) is allowed to define those two bytes (and the rest of the packet!) any way they like. Rumor had it that Xerox had registered OUI==0 and had pronounced that their definition for those two bytes would be identical to Ethernet type codes ("SNAPpy Xerox"). Whether or not that rumor was correct, the present de-facto industry and Internet standard is that "OUI==0" means the next two bytes are an Ethernet "type" which describes the rest of the packet, thus allowing good old Ethernet packets to be encapsulated in 802.2 LLC frames. See RFC 1042 and RFC 1103 for a more formal description of the LLC/SNAP/OUI==0 encapsulation. Note that the use of "Ethernet types" for those two bytes applies *ONLY* to OUI==0. Other organizations may (and have?) used SNAP (with OUI != 0) for their own proprietary purposes... +--------------- | 3) What SNAPs have been defined to date? +--------------- Besides OUI==0 (RFC 1042/1103), who knows...? Try the IEEE. -Rob ----- Rob Warnock, MS-1L/515 rpw3@sgi.com rpw3@pei.com Silicon Graphics, Inc. (415)335-1673 Protocol Engines, Inc. 2011 N. Shoreline Blvd. Mountain View, CA 94039-7311