In 2005 speed wasn't the problem but redundancy was difficult to design in ethernet. I'm not a tokenring fan but the relianility of FDD- was impressive.
Verzonden vanaf mijn draadloze BlackBerry -toestel
From: Jason Stevens <neozeed at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Tue, 12 Jul 2011 17:39:45 -0400
To: <hecnet at update.uu.se>
ReplyTo: hecnet at Update.UU.SE
Subject: Re: [HECnet] Towards the Mouth of Madness....
I thought all the new ethernet stuff does jumbo frames for those that like that kind of thing anyways....
Not true, not unless the implementation is crummy. Any halfway decent fast Ethernet host will run at wire speed, and the difference between 1500 and 4460 MTU isn't enough to amount to very much.
paul
Yes, absolutely! The bandwith was 98% of the nominal value. Ethernet starts diminshing after 60%.
I did extensive testing of FDDI back in early -90's and put as much load on it as I could with about a dozen Alphas of the time and found out that it easily outperformed Ethernet.
FDDI gear were more expensive than Ethernet, but were more useful.
FDDI was really outperformed by GbE as the multiplexed 100Mbit/s wasn't available on so many platforms.
Kari
On 13.7.2011 0:29, hvlems at zonnet.nl wrote:
It took HP until 2006 to replace the FDDI lan with ethernet technolgy.
The fault tolerance of FDDI and the build quality of DEC's gigaswitch
products.
The bandwidth of FDDI is a lot better than fast ethernet.
Hans
Verzonden vanaf mijn draadloze BlackBerry -toestel
------------------------------------------------------------------------
*From: * Joe Ferraro <jferraro at gmail.com>
*Sender: * owner-hecnet at Update.UU.SE
*Date: *Tue, 12 Jul 2011 17:12:02 -0400
*To: *<hecnet at update.uu.se>
*ReplyTo: * hecnet at Update.UU.SE
*Subject: *Re: [HECnet] Towards the Mouth of Madness....
Off topic... I received a page, a few weeks prior, on a machine that was
not pinging. Turns out, it was one of a few old NOVA class boxes we
still have at my work, using FDDI for connectivity. Fortunately, a
disconnect / reconnect brought the ring back online; I was scared (and a
bit excited in a strange way) for a few moments that I was going to have
to do some extensive troubleshooting... FDDI still lives.
On Tue, Jul 12, 2011 at 4:27 PM, H Vlems <hvlems at zonnet.nl
<mailto:hvlems at zonnet.nl>> wrote:
__
Remember what I wrote: this happened nearly two decades ago.____
IP is the protocol that survived and most people aren t even aware
what happened on local area networks before, say,1998.____
I worked for Fuji, photosensitive films, paper and offset printing
products.____
Most of the IT equipment was made by DEC: PDP-11 s (/44, /84, /93,
/24, /73 and /23), VAXes, an IBM mainframe (4081) and PC s.____
And lots of other gear, most of it in the research lab. A Motorola
box that ran Motorola Unix, and an RS/6000 under AIX 2.4 (?).____
The lingua franca was DECnet and LAT. No IP, though some PC s used
Novell and SNA over tokenring to make terminal emulation to the
mainframe possible.____
No IP. Sounds weird in today s world but DECnet eventually
connected everything. We got a **very** early Cisco router that did
level 1____
DECnet routing between the corporate ethernet and the finance dept
token ring. Another (DEC) box that routed DECnet over Datanet/1
(that s X25 in Europe IIRC). The mainframe used an SNA/DECnet
gateway (the big channel attached box).____
The RS/6000 and the Motorola systems also ran DECnet, endnode only.____
To make this a little interesting we ran the first FDDI network in
the Netherlands.____
Trouble shooting wasn t always easy, especially if the SNA/DECnet
gateway was involved!____
Hans____
__ __
------------------------------------------------------------------------
*Van:*owner-__hecnet at Update.UU.SE <mailto:hecnet at Update.UU.SE>__
[mailto:owner- <mailto:owner->__hecnet at Update.UU.SE
<mailto:hecnet at Update.UU.SE>__] *Namens *Jason Stevens
*Verzonden:* dinsdag, juli 2011 21:10
*Aan:* hecnet at update.uu.se <mailto:hecnet at update.uu.se>
*Onderwerp:* Re: [HECnet] Towards the Mouth of Madness....____
__ __
AIX and decnet? now that'd be ... non conformist & fun! ____
------------------------------------------------------------------------
Geen virus gevonden in dit bericht.
Gecontroleerd door AVG - www.avg.com <http://www.avg.com>
Versie: 10.0.1388 / Virusdatabase: 1516/3760 - datum van uitgifte:
07/12/11____
Metro ethernet & the move to VIOP on the carrier side killed ATM. But there was a big push from the mid 90's until 2005 or so from what I noticed. LANE was so fun to configure (LECS/LES/BUS/LEC) along with such friendly addressesing... I can't say it'll be missed. But you could get quite creative with it.
What really killed ATM IMHO was our carrier (Bell South) would charge the same price for CBR-RT VBR-RT pvp's as what you would pay for a T1. So they made the technology moot. Esp when you have AT&T G3's (Lucent Avaya etc..) which only deal in T1's it all became a lost cause.
Not to mention cheap & ATM certainly didn't go hand in hand. Esp with the Cisco LS1010.
On Tue, Jul 12, 2011 at 5:40 PM, <hvlems at zonnet.nl> wrote:
FDDI was great in 1991. It could easily span a large multi building site (> 4 km^2) and with a bandwidth that was better than that of fast ethernet. That didn't exist at the time, 100VG was on the drawing boards. The other alternative was 155 Mb/s ATM but that technology never really took off (in Europe, dunno about the US).
Verzonden vanaf mijn draadloze BlackBerry -toestel
From: Jason Stevens <neozeed at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Tue, 12 Jul 2011 17:31:03 -0400
To: <hecnet at update.uu.se>
ReplyTo: hecnet at Update.UU.SE
Subject: Re: [HECnet] Towards the Mouth of Madness....
FEC (fast ether channel) killed any hopes of FDDI, but it was more so people in certain places that bought into the hopes and dreams of FDDI....
It was also my understanding you needed licenses to make FDDI gear while Ethernet was/is free.
On Tue, Jul 12, 2011 at 5:29 PM, <hvlems at zonnet.nl> wrote:
It took HP until 2006 to replace the FDDI lan with ethernet technolgy. The fault tolerance of FDDI and the build quality of DEC's gigaswitch products.
The bandwidth of FDDI is a lot better than fast ethernet.
Hans
Verzonden vanaf mijn draadloze BlackBerry -toestel
From: Joe Ferraro <jferraro at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Tue, 12 Jul 2011 17:12:02 -0400
To: <hecnet at update.uu.se>
ReplyTo: hecnet at Update.UU.SE
Subject: Re: [HECnet] Towards the Mouth of Madness....
Off topic... I received a page, a few weeks prior, on a machine that was not pinging. Turns out, it was one of a few old NOVA class boxes we still have at my work, using FDDI for connectivity. Fortunately, a disconnect / reconnect brought the ring back online; I was scared (and a bit excited in a strange way) for a few moments that I was going to have to do some extensive troubleshooting... FDDI still lives.
On Tue, Jul 12, 2011 at 4:27 PM, H Vlems <hvlems at zonnet.nl> wrote:
Remember what I wrote: this happened nearly two decades ago.
IP is the protocol that survived and most people aren t even aware what happened on local area networks before, say,1998.
I worked for Fuji, photosensitive films, paper and offset printing products.
Most of the IT equipment was made by DEC: PDP-11 s (/44, /84, /93, /24, /73 and /23), VAXes, an IBM mainframe (4081) and PC s.
And lots of other gear, most of it in the research lab. A Motorola box that ran Motorola Unix, and an RS/6000 under AIX 2.4 (?).
The lingua franca was DECnet and LAT. No IP, though some PC s used Novell and SNA over tokenring to make terminal emulation to the mainframe possible.
No IP. Sounds weird in today s world but DECnet eventually connected everything. We got a *very* early Cisco router that did level 1
DECnet routing between the corporate ethernet and the finance dept token ring. Another (DEC) box that routed DECnet over Datanet/1 (that s X25 in Europe IIRC). The mainframe used an SNA/DECnet gateway (the big channel attached box).
The RS/6000 and the Motorola systems also ran DECnet, endnode only.
To make this a little interesting we ran the first FDDI network in the Netherlands.
Trouble shooting wasn t always easy, especially if the SNA/DECnet gateway was involved!
Hans
Van: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] Namens Jason Stevens Verzonden: dinsdag, juli 2011 21:10 Aan: hecnet at update.uu.se Onderwerp: Re: [HECnet] Towards the Mouth of Madness....
AIX and decnet? now that'd be ... non conformist & fun!
Geen virus gevonden in dit bericht. Gecontroleerd door AVG - www.avg.com Versie: 10.0.1388 / Virusdatabase: 1516/3760 - datum van uitgifte: 07/12/11
Paul. In 1991 there was no fast ethernet as we know it today. Ethernet over glassfiber was proprietary and by and large restricted to 10 Mb/s.
The problems with the redesign had tot do with emulating FDDI's built in redundancy with ethernet.
------Origineel bericht------
Van: Paul Koning
Afzender: owner-hecnet at Update.UU.SE
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: Re: [HECnet] Towards the Mouth of Madness....
Verzonden: 12 juli 2011 23:38
On Jul 12, 2011, at 5:29 PM, <hvlems at zonnet.nl> <hvlems at zonnet.nl> wrote:
It took HP until 2006 to replace the FDDI lan with ethernet technolgy. The fault tolerance of FDDI and the build quality of DEC's gigaswitch products.
The bandwidth of FDDI is a lot better than fast ethernet.
Not true, not unless the implementation is crummy. Any halfway decent fast Ethernet host will run at wire speed, and the difference between 1500 and 4460 MTU isn't enough to amount to very much.
paul
Verzonden vanaf mijn draadloze BlackBerry -toestel
On 2011-07-12 23:14, Joe Ferraro wrote:
While we are on the topic, does anyone have any experience with TOPS-20
/ DECNET? I've haven't had time to mess with it of late, but I was
unsuccessful in a previous evening's attempt or two.
Well, doh...! :-)
.ncp tell sol sho exec cha
Node characteristics as of 12-JUL-11 23:38:44
Executor node = 59.10 (SOL)
Identification = Systems Concepts SF CA USA - SC30M - DN-20 4.0, Management v
ersion = 4.0.0
Loop count = 1, Loop length = 127
Loop with = Mixed, Incoming timer = 30
Outgoing timer = 60, NSP version = 4.0.0
Maximum links = 65535, Delay factor = 48
Delay weight = 10
Inactivity timer = 120, Retransmit factor = 10
Routing version = 2.0.0, Type = Routing IV
Routing timer = 600
Broadcast routing timer = 40, Maximum address = 1023
Maximum circuits = 20
Maximum cost = 100
Maximum hops = 16, Maximum visits = 20
Maximum broadcast endnodes = 64
Maximum broadcast routers = 32
Maximum buffers = 80, Buffer size = 576
Segment buffer size = 576
.
So, yes, there are people around... Like I said - not everyone on this list is even running a VMS system...
Johnny
On Tue, Jul 12, 2011 at 5:11 PM, Mark Benson <md.benson at gmail.com
<mailto:md.benson at gmail.com>> wrote:
On 12 Jul 2011, at 21:48, hvlems at zonnet.nl <mailto:hvlems at zonnet.nl>
wrote:
> DECnet/AIX was a 3rd party effort IIRC
> I haven't searched the internet yet but who knows what is available.
> Perhaps the opensource decnet kit may be portable.
I know there's a DECNet kit for Linux out there somewhere. Not sure
about anything else. Nothing crops up on a quick Google search that
isn't dated early 90s.
> I wasn't aware that AIX was so hobbyist friendly. The hardware is
vey unfamiliar for me though.
The hardware is generally pretty unique (save for it's similarity in
patches to PPC Macs) - all IBM stuff generally is. The OS is
proprietary but you can pick up media kits off eBay for not a lot.
Also IBM documentation, certainly for RS/6000 / pSeries hardware and
AIX, is readily available online without even needing a login.
--
Mark Benson
My Blog:
<http://markbenson.org/blog>
Follow me on Twitter:
http://twitter.com/mdbenson
"Never send a human to do a machine's job..."
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
FDDI was great in 1991. It could easily span a large multi building site (> 4 km^2) and with a bandwidth that was better than that of fast ethernet. That didn't exist at the time, 100VG was on the drawing boards. The other alternative was 155 Mb/s ATM but that technology never really took off (in Europe, dunno about the US).
Verzonden vanaf mijn draadloze BlackBerry -toestel
From: Jason Stevens <neozeed at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Tue, 12 Jul 2011 17:31:03 -0400
To: <hecnet at update.uu.se>
ReplyTo: hecnet at Update.UU.SE
Subject: Re: [HECnet] Towards the Mouth of Madness....
FEC (fast ether channel) killed any hopes of FDDI, but it was more so people in certain places that bought into the hopes and dreams of FDDI....
It was also my understanding you needed licenses to make FDDI gear while Ethernet was/is free.
On Tue, Jul 12, 2011 at 5:29 PM, <hvlems at zonnet.nl> wrote:
It took HP until 2006 to replace the FDDI lan with ethernet technolgy. The fault tolerance of FDDI and the build quality of DEC's gigaswitch products.
The bandwidth of FDDI is a lot better than fast ethernet.
Hans
Verzonden vanaf mijn draadloze BlackBerry -toestel
From: Joe Ferraro <jferraro at gmail.com>
Sender: owner-hecnet at Update.UU.SE
Date: Tue, 12 Jul 2011 17:12:02 -0400
To: <hecnet at update.uu.se>
ReplyTo: hecnet at Update.UU.SE
Subject: Re: [HECnet] Towards the Mouth of Madness....
Off topic... I received a page, a few weeks prior, on a machine that was not pinging. Turns out, it was one of a few old NOVA class boxes we still have at my work, using FDDI for connectivity. Fortunately, a disconnect / reconnect brought the ring back online; I was scared (and a bit excited in a strange way) for a few moments that I was going to have to do some extensive troubleshooting... FDDI still lives.
On Tue, Jul 12, 2011 at 4:27 PM, H Vlems <hvlems at zonnet.nl> wrote:
Remember what I wrote: this happened nearly two decades ago.
IP is the protocol that survived and most people aren t even aware what happened on local area networks before, say,1998.
I worked for Fuji, photosensitive films, paper and offset printing products.
Most of the IT equipment was made by DEC: PDP-11 s (/44, /84, /93, /24, /73 and /23), VAXes, an IBM mainframe (4081) and PC s.
And lots of other gear, most of it in the research lab. A Motorola box that ran Motorola Unix, and an RS/6000 under AIX 2.4 (?).
The lingua franca was DECnet and LAT. No IP, though some PC s used Novell and SNA over tokenring to make terminal emulation to the mainframe possible.
No IP. Sounds weird in today s world but DECnet eventually connected everything. We got a *very* early Cisco router that did level 1
DECnet routing between the corporate ethernet and the finance dept token ring. Another (DEC) box that routed DECnet over Datanet/1 (that s X25 in Europe IIRC). The mainframe used an SNA/DECnet gateway (the big channel attached box).
The RS/6000 and the Motorola systems also ran DECnet, endnode only.
To make this a little interesting we ran the first FDDI network in the Netherlands.
Trouble shooting wasn t always easy, especially if the SNA/DECnet gateway was involved!
Hans
Van: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] Namens Jason Stevens Verzonden: dinsdag, juli 2011 21:10 Aan: hecnet at update.uu.se Onderwerp: Re: [HECnet] Towards the Mouth of Madness....
AIX and decnet? now that'd be ... non conformist & fun!
Geen virus gevonden in dit bericht. Gecontroleerd door AVG - www.avg.com Versie: 10.0.1388 / Virusdatabase: 1516/3760 - datum van uitgifte: 07/12/11
On Jul 12, 2011, at 5:31 PM, Jason Stevens wrote:
FEC (fast ether channel) killed any hopes of FDDI, but it was more so people in certain places that bought into the hopes and dreams of FDDI....
It was also my understanding you needed licenses to make FDDI gear while Ethernet was/is free.
No. But the chips were far more expensive and you needed more of them to make a NIC. While it was in theory possible to make a single chip FDDI NIC, I don't think that was ever produced. And even if it were, the volumes would be far lower which means the price is far higher.
paul
I thought all the new ethernet stuff does jumbo frames for those that like that kind of thing anyways....
Not true, not unless the implementation is crummy. Any halfway decent fast Ethernet host will run at wire speed, and the difference between 1500 and 4460 MTU isn't enough to amount to very much.
paul
On Jul 12, 2011, at 5:29 PM, <hvlems at zonnet.nl> <hvlems at zonnet.nl> wrote:
It took HP until 2006 to replace the FDDI lan with ethernet technolgy. The fault tolerance of FDDI and the build quality of DEC's gigaswitch products.
The bandwidth of FDDI is a lot better than fast ethernet.
Not true, not unless the implementation is crummy. Any halfway decent fast Ethernet host will run at wire speed, and the difference between 1500 and 4460 MTU isn't enough to amount to very much.
paul
While we are on the topic, does anyone have any experience with TOPS-20 / DECNET? I've haven't had time to mess with it of late, but I was unsuccessful in a previous evening's attempt or two.
I've had mixed luck with it. I did manage to get it up and running on my Panda distribution, but then I changed DECnet addresses, and my attempt to get it working again didn't go well. I believe that NCP continued to complain that my DECnet address was invalid (perhaps saying that the number was greater than 63, but it wasn't).
I don't have the system up right now and I'm remote at the moment, but if there are any specifics I can try to recall for you, I'll do what I can.
I will say that I modified the hardware address of the Unix (Solaris, in my case) network interface and also in the Panda configuration for the NI device to reflect the appropriate AA-00-04-00-xx-xx value. The place where I got stuck was in the NCP configuration in whichever <SYSTEM> .CMD file it was that the system uses at boot time.
--Marc
