Title: | DEChub/HUBwatch/PROBEwatch CONFERENCE |
Notice: | Firmware -2, Doc -3, Power -4, HW kits -5, firm load -6&7 |
Moderator: | NETCAD::COLELLA DT |
Created: | Wed Nov 13 1991 |
Last Modified: | Fri Jun 06 1997 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 4455 |
Total number of notes: | 16761 |
From: US1RMC::"[email protected]" "GIGAswitch Notes Daemon - entry by [email protected]" 15-AUG-1994 09:42:14.86 To: <cern::jrs> CC: Subj: Making ringmaps without ELMS <<< AXPVMS::SYS$SYSDEVICE:[NOTES$LIBRARY]GIGASWITCH.NOTE;1 >>> -< CERN GIGAswitch Notes Conference >- ================================================================================ Note 118.0 Making ringmaps without ELMS No replies AXPVMS::GIGASWITCH 39 lines 15-AUG-1994 15:23 -------------------------------------------------------------------------------- Date Of Receipt: 15-AUG-1994 15:06:55.77 From: DXMINT::"[email protected]" To: axpvms::GIGAswitch CC: Subj: Making ringmaps without ELMS With the advent of the GIGAswitch and the demise of ELMS to make ringmaps, we need to have other means to produce maps of the rings connected to the GIGAswitch and/or behind IP routers. Although it is true that the fddimib gives values for "fddimibMACUpstreamNbr" and "fddimibMACDownstreamNbr", many systems on FDDI have not implemented these and will probably never have it. The only solution to this problem seems to be a "proxy agent" that is addressed with SNMP and that collects the relevant information via SMT. So our question is if you have envisaged a solution to this problem Moreover, a solution implemented in the GIGAswitch would be really good (leaving the problem to be solved for a ring behind a router). I am not aware of a "standard" solution for this, although I know at least one manufacturer who has got it (proprietary solution). Your advice, ideas and plans will be greatly appreciated. ================== RFC 822 Headers ================== Received: by dxcoms.cern.ch (5.65/DEC-Ultrix/4.3) id AA14321; Mon, 15 Aug 1994 15:07:25 +0200 Message-Id: <[email protected]> Date: Mon, 15 Aug 1994 15:07:24 +0200 (MET DST) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 916 % ====== Internet headers and postmarks (see DECWRL::GATEWAY.DOC) ====== % Received: from inet-gw-3.pa.dec.com by us1rmc.bb.dec.com (5.65/rmc-22feb94) id AA05046; Mon, 15 Aug 94 09:41:34 -040 % Received: from dxmint.cern.ch by inet-gw-3.pa.dec.com (5.65/10Aug94) id AA26528; Mon, 15 Aug 94 06:25:53 -070 % Received: from AXPVMS.DECnet MAIL11D_V3 by dxmint.cern.ch (5.65/DEC-Ultrix/4.3) id AA17868; Mon, 15 Aug 1994 15:24:23 +020 % Date: Mon, 15 Aug 1994 15:24:23 +0200 % Message-Id: <[email protected]> % From: [email protected] (GIGAswitch Notes Daemon - entry by [email protected]) % X-Vms-To: DXMINT::[email protected],DXMINT::[email protected],DXMINT::[email protected],DXMINT::[email protected],DXMINT::[email protected],DXMINT::jmj@dxcoms % Subject: Making ringmaps without ELMS % X-Mail11-Ostype: VAX/VMS % Apparently-To: <[email protected]> % Apparently-To: <npss::mdlyons> % Apparently-To: <npss::renda> % Apparently-To: <school::gstest> % Apparently-To: <cern::bothner> % Apparently-To: <cern::jrs>
T.R | Title | User | Personal Name | Date | Lines |
---|---|---|---|---|---|
1355.1 | Look in version 4 | SLINK::HOOD | I'd rather be at the Kennebec | Fri Aug 26 1994 17:57 | 4 |
Ring maps (both FDDI and Token Ring) are planned for HUBwatch V4.0. Tom Hood HUBwatch | |||||
1355.2 | SNMP/SMT Gateway. | NACAD::GALLAGHER | Fri Aug 26 1994 18:30 | 17 | |
Yes, we've recognized the problem and have begun implementing a solution. An SMT (FDDI Station Management) gateway is planed for the DECconcentrator900MX. The gateway works like the "proxy agent" you describe. The gateway can be easily ported to the DECbridge900 series products as well, though that activity is not scheduled. No such activity is scheduled for the GigaSwitch either. >I am not aware of a "standard" solution for this, although I know >at least one manufacturer who has got it (proprietary solution). I'm not aware of a standard solution either. I'm interested in any details you have on the other manufacturer's solution. Please contact me if you have any details. -Shawn | |||||
1355.3 | NPSS::MDLYONS | Michael D. Lyons - Young enough and dumb enough | Mon Aug 29 1994 15:02 | 6 | |
How would this be implemented in HUBwatch without the proxy SMT agent? ...Thanks MDL | |||||
1355.4 | NACAD::GALLAGHER | Mon Aug 29 1994 15:39 | 16 | ||
I'm not sure I understand the question. If you're questioning the difference between an "SMT Gateway" and an "SMT/SNMP Proxy Agent" - there's very little difference. A HUBwatch application like FDDI ring map is invoked. HUBwatch packages up an SMT frame, with the appropriate padding for source informations. It then issues a set to the device, setting an object like, say, "smtGwRequest- Frame". The device takes the data (the SMT request frame), overwrites the source information with its source address, etc., and sends the frame. The device puts the response SMT frame(s) into "smtGwRespFrame". HUBwatch reads "smtGwRespFrame", interprets the results, updates the map, and repeats until the ring configuration is known. (Of course, other objects are involved to provide synchronization, make the solution more generic, etc.) -Shawn | |||||
1355.5 | DEChub - based FDDI maps! | SLINK::HOOD | I'd rather be at the Kennebec | Mon Aug 29 1994 15:50 | 5 |
Sorry for the confusion. The FDDI ring map that we hope will be in HUBwatch V4.0 will require a DEChub-based ring. GIGAswitch-based rings (without a DEChub connection) will not be supported Tom Hood | |||||
1355.6 | NPSS::MDLYONS | Michael D. Lyons - Young enough and dumb enough | Mon Aug 29 1994 16:56 | 1 | |
Re: .5 Thanks, that's what I was asking... | |||||
1355.7 | Only one gateway device needed on ring | NACAD2::HAROKOPUS | Mon Aug 29 1994 18:59 | 7 | |
Note that you will only need one DEChub FDDI module that supports the SNMP-SMT gateway. The rest of the ring can consist of any vendors FDDI devices as long as the conform to the ANSI standard. Regards, Bob |