callahan@netman.DEC (JOE CALLAHAN TWO/E17 DTN247-2843) (06/27/85)
[]
seismo!camelot!schoff asked for info regarding DEC products to put
ETHERNET on a broadband cable plant.
Digital does offer a set of products to permit ETHERNET operation
over broadband cable systems. They are alternatives to the baseband
transceiver but look just like it to all the other ETHERNET
devices, e.g. DEUNA, DELNI, etc.
---------------------------------
PRODUCT DESCRIPTIONS
Broadband Ethernet Transceiver
The Broadband Ethernet Transceiver (DECOM) is the device that provides
physical and electrical interface to the broadband coaxial cable. The
transceiver transmits signals onto the coaxial cable, receives signals from
the cable, and detects any message collisions that may occur.
Single Cable Broadband Ethernet Transceiver (DECOM-BA/BB)
The single cable Broadband Ethernet Transceiver transmits in the frequency
band of 54-to-72 megahertz (MHz) and receives at 210.25-to-228.25 MHz. This
transceiver requires a Frequency Translator (DEFTR) at the network headend
to convert the Broadband Ethernet signals from their transmit to their
receive frequencies. One DEFTR is required per network.
Dual Cable Broadband Ethernet Transceiver (DECOM-AA/AB)
The dual cable Broadband Ethernet Transceiver transmits and receives the
Ethernet signals at the same frequencies, in the band from 54-to-72 MHz.
This model does not require a Frequency Translator (DEFTR) at the network
headend.
-------------------------------
I don't have current info on prices and availability. For that
contact the Digital sales office nearest you. If the person you
reach has trouble with your questions, ask to talk to someone from
the "Network Sales Team".
(Usual disclaimers re: these being my own and not my employer's opinions,
but in this case I'd better be pretty
close to being "right")
/s/ Joe Callahan (... then is Real Intelligence that which has not
(yet) been reduced to a set of rules? )
(DEC Easynet) NETMAN::CALLAHAN
(UUCP) {decvax, ucbvax, allegra}!decwrl!dec-rhea!dec-netman!callahan
(ARPA) callahan%netman.DEC@decwrl.ARPA
(Telephone) 617-858-2843 root@bu-cs.UUCP (Barry Shein) (06/29/85)
As another thought, at Boston University we have solved the problem with our driver (a major adaptation of the UCB driver) which talks over our Ungermann/Bass via a DR11-W attached to an NIU-150 (U/B broadband modem.) It provides the exact same functions as an ethernet interface (even ARP and Broadcasts) to the higher layers. It goes in as just another interface under 4.2 TCP/IP and some of our machines have both the broadband and ethernet so they act as gateways. It all works fine, has been in production here for several months under both 4.2 (VAX) and VMS (Wollongong's TCP/IP), Suns soon. -Barry Shein, Boston University
schoff@rpics.UUCP (Marty Schoffstall) (06/29/85)
> > PRODUCT DESCRIPTIONS > > Broadband Ethernet Transceiver > The Broadband Ethernet Transceiver (DECOM) is the device that provides > physical and electrical interface to the broadband coaxial cable. The > transceiver transmits signals onto the coaxial cable, receives signals from > the cable, and detects any message collisions that may occur. According to DEC this costs $4500. > > Single Cable Broadband Ethernet Transceiver (DECOM-BA/BB) > The single cable Broadband Ethernet Transceiver transmits in the frequency > band of 54-to-72 megahertz (MHz) and receives at 210.25-to-228.25 MHz. This > transceiver requires a Frequency Translator (DEFTR) at the network headend > to convert the Broadband Ethernet signals from their transmit to their > receive frequencies. One DEFTR is required per network. > According to DEC this costs $4250. -- marty schoff%rpi@csnet-relay ARPA schoff@rpi CSNET seismo!rpics!schoff UUCP
mccallum@opus.UUCP (Doug McCallum) (06/30/85)
> As another thought, at Boston University we have solved the problem with > our driver (a major adaptation of the UCB driver) which talks over our > Ungermann/Bass via a DR11-W attached to an NIU-150 (U/B broadband > modem.) It provides the exact same functions as an ethernet interface > (even ARP and Broadcasts) to the higher layers. > -Barry Shein, Boston University The advantage of the ethernet on broadband is that it works with existing ethernet controllers and drop cables. That means it should work with non-DEC equipment and no new hardware has to be added to use it if the system already has ethernet support. Disadvantages are cost and the number of channels it takes on the broadband system (equivalent to 3 CATV channels). Doug McCallum NBI, Inc. {allegra, ucbvax, ut-sally}!nbires!mccallum
root@bu-cs.UUCP (Barry Shein) (07/02/85)
Re: DR11-W--U/B NIU-150 --Broadband (Ungermann/Bass Broadband driver for 4.2,VMS at Boston University.) Just to clarify (a comment made by a poster indicated some confusion): The reason I entered this into the discussion was because our broadband interfaces *do* act as ethernet/broadband bridges, exactly like you would hope. We have two 4.2 vaxes with both an ethernet (DEUNA) and U/B Broadband interface which act as transparent gateways for other non-broadband hosts, for example: Eng Vax IBM3081 C.S. Student Vax C.S. Research Vax DEC2060 broadband | street | ethernet ethernet only ethernet+broadband ethernet only only only Every system in the picture can TELNET, FTP, SMTP etc to every other system in the picture as if they were on one ethernet (although because they form logically distinct subnets broadcasts have to be forwarded, like two ethernets.) I hope this clarifies something. No, does not require a DEC system, we are soon to move it to a SUN, any 4.2bsd system with sources and a DR11-W or DR11-B compatible interface would do (MultiBus,VMEbus DR11-W emulators exist.) You only need one such system per ethernet anyhow which acts as a gate. It only takes one channel on the broadband which is important to us. We will look into other interfaces such as RS232 and V.35 also to lessen the DR11-W constraint (U/B would probably have some ideas, they make several interfaces.) Yes, I would like a black box that just snapped onto an ethernet cable and broadband tap and magically forwarded to another, remote box with the same connections, but I don't like the idea of 2 or 3 dedicated channels to do it. I also wonder exactly what gets forwarded (every packet? the DEC box certainly doesn't read IP packets.) I guess for now I am using a couple of Vaxes as this 'black box'. I could imagine buying a cheap 68K 4.2 box, adding a parallel and ethernet interface and configuring in the driver and our campus' routing tables and thus acheiving this, tho probably a little expensive (well, a bit less than $10,000 for sure.) And hell, a few people could log into it also, or it could be expanded to be a print server or some such. (note that we have yet to see a performance problem, packets get forwarded thru 4.2 very cheaply.) Nah, we looked at the DEC broadband box as we were writing our own. Unless you are locked into DecNet we didn't think it was a very good solution at all, tho I will admit the choices are limited. -Barry Shein, Boston University
mccallum@opus.UUCP (Doug McCallum) (07/03/85)
> Yes, I would like a black box that just snapped onto an ethernet cable > and broadband tap and magically forwarded to another, remote box with > the same connections, but I don't like the idea of 2 or 3 dedicated > channels to do it. I also wonder exactly what gets forwarded (every > packet? the DEC box certainly doesn't read IP packets.) I guess for now > I am using a couple of Vaxes as this 'black box'. The DEC ethernet on broadband isn't a black box that forward packets, although you could use the Vitalink box using baseband on oneside and broadband on the other. The DEC product is a broadband tranceiver that connects to an ethetnet/802.3 controller on whatever type of system you have. It even uses the same drop cable as far as I can tell. What gets put on the coax is exactly the same bits that would go on the baseband coax. It really is a broadband implementation of 802.3 CSMA/CD at 10Mb/s. It is being considered as a new family member to the IEEE 802.3 family of access methods. Given this form, there are no protocol dependencies. You could run a mixture of TCP/IP, XNS, DECNET, etc. over the same cable as you can do now with the baseband implementation. It requires no new hardware or software other than what is needed to connect to the cable (ie tranceivers and a headend). I intend this article to be informational only and not specifying a preference. Both the U-B approach and the DEC approach have advantages and disadvantages. Just to throw more confusion our, the Sytek network which IBM is using as their PC network is a broadband network using an 802.3 compatible packet format. It only uses 1 CATV channel but operates at lower speed. It needs a new controller and tap since it uses a different scheme for encoding the signal.
mike@brl-sem.ARPA (Michael John Muuss <mike>) (07/06/85)
We do the same thing, only using a Proteon V2LNI Ringnet ("Pronet")
at 10 Mbps over fiber optic links. This forms our "Campus Area Net"
(CAN?), connecting all our LANs together.
We have some 80 Mbps Pronet on order; then things will get really fast!
-Mikesteveg@hammer.UUCP (Steve Glaser) (07/10/85)
>Yes, I would like a black box that just snapped onto an ethernet cable >and broadband tap and magically forwarded to another, remote box with >the same connections, but I don't like the idea of 2 or 3 dedicated >channels to do it. I also wonder exactly what gets forwarded (every >packet? the DEC box certainly doesn't read IP packets.) I guess for now >I am using a couple of Vaxes as this 'black box'. I think Bridge makes a box that connects 2 ethernets over a point to point link (V.35 or whatever at any data rate you care about). It works directly on the Ethernet layer, making no assumptions about what's on top of that (IP, DECNET, XNS, etc.). With a simple RF modem, you can easily get this "point-to-point" link onto a broadband cable. It essentially adapts itself to the ethernet addresses of the stuff on either end and uses that to help it decide what needs forwarding across the link. Given that there are only about 100 hosts per Ethernet the tables aren't that big a deal, especially if you don't need to generate them. I suppose this only really works on protocols like TCP where retransmission times adapt to the network rather than "knowing" what an Ethernet can do (most TCP implementations seem to do this). Also, I don't know what this box does with broadcast packets - I suppose they get forwarded (otherwise ARP etc. wouldn't work). I expect that this solution may break (or at least get bogged down) when folks really start using multicast packets on Ethernet. It certainly will have problems if there is a large speed disparity between 10Mbps and the point-to-point link. Steve Glaser Tektronix Inc. /* usual disclaimer goes here */
sjl@amdahl.UUCP (Steve Langdon) (07/11/85)
> I think Bridge makes a box that connects 2 ethernets over a point to > point link (V.35 or whatever at any data rate you care about). It > works directly on the Ethernet layer, making no assumptions about > what's on top of that (IP, DECNET, XNS, etc.). With a simple RF modem, > you can easily get this "point-to-point" link onto a broadband cable. I believe that what you are referring to is actually a product called TransLan (sp?) sold by Vitalink the satellite people. They developed it in conjunction with DEC and they use the Bridge hardware/software which they buy on an OEM basis. The actual software which does the adaptive forwarding was developed by Vitalink. The forwarding is not limited to use on a local point-to-point link, but will also work using a satellite channel operating in broadcast mode. Thus many Ethernets spread over the country can be made to look (logically) like one LAN. I'm sure that they would be more than delighted to tell you how it works and the results of their performance tests. They are in Mountain View, California (near Sun, 3Com, Bridge, etc.) and their phone number is (415)968-5465. As usual I have nothing to do with Vitalink, I just think it is a clever product. -- Stephen J. Langdon ...!{ihnp4,hplabs,sun,nsc}!amdahl!sjl [ The article above is not an official statement from any organization in the known universe. ]
jbn@wdl1.UUCP (08/05/85)
Bridge makes IP and XNS gateways, with real routing algorithms; we will be evaluating the IP model next month, and will report the results. The Vitalink-modified Bridge box is a nice idea, but one of the byproducts of their concept is that all the broadcast traffic goes to everybody in the whole system, which can be a problem. Vitalink assumes that you have lots of bandwidth between the sites. Since Vitalink's real business is satellite ground stations, this works out quite well for them. But it's not an optimal use of bandwidth. John Nagle
bruce@ssc-vax.UUCP (Bruce Stock) (11/04/85)
Is anyone making use of the DEC DECOM (or MA/COM) broadband ethernet transciever? Any nasty surprises? How does it work out in a production environment? Insight will be appreciated, Bruce Stock Boeing Aerospace