---------------------------------------------------------------------------- NOV-FDDI.DOC -- 19970121 -- Email thread on Fiber Data Distributed Interface ---------------------------------------------------------------------------- Feel free to add or edit this document and then email it back to faq@jelyon.com Date: Thu, 20 Jun 1996 15:14:59 -0600 From: Joe Doupnik Subject: Re: Better LAN configuration >I had not read any discussion on performance and throughput for the >following network configuration/connetions. If anyone had any test or >insights please feel free to share with me. Basically, I am thinking of >setting a LAN with one of two method. > >1) Using a 100Mbps connections to the server and rest of the worstations >are connected using 10Mbps to the ethernet switch hub. > > Netware server > || > || 100Mbps(1 line) > || > ---------------- > | switch hub | > ---------------- > / | | | \ 10Mbps > / | | | \ > workstations and/or regular hubs. > >2) Using 4 network cards in the server and connect it to the switch hub. >All connections are 10Mbps. Lines to the servers are full parity. I guess you meant "full duplex" > Netware server > | | | | > | | | | 10Mbps(4 lines) > | | | | > ---------------- > | switch hub | > ---------------- > / | | | \ 10Mbps > / | | | \ > workstations and/or regular hubs. > >Many users on my LAN use Microsoft Access Database that is around 10M in >size. As well, I also have a sql database running with over 50M of data. >As well, there is also a tcp/ip router coming in which is not shown. ------------ The top configuration is the easiest to achieve. The bottom one should have somewhat better throughput today (not necessarily tomorrow) because 100Mbps Ethernet still has performance problems in the drivers. On the other hand, finding a hub with four full duplex ports might be difficult (anyone know?), and you would depend on load balancing software such as say NLSP to run the boards in parallel like that. Load balancing with NLSP handles IPX traffic ok, but doesn't deal with IP traffic, so there's a minus to consider. By way of comparison, Win95 loading is about 7.5MB here. I've been contemplating the top configuration to replace a three wire spread, but the funds just aren't there right now. If I had the money I would use FDDI between server and hub to get almost all of 100Mbps rather than just 20Mbps or so with today's software; but FDDI == $$$$. Joe D. ------------------------------ Date: Mon, 24 Jun 1996 13:35:52 +0100 From: Richard Letts Subject: Re: FDDI vs. 100 MB ethernet >Besides cost, is there any advantage to fast (and switched) ethernet over >FDDI? > >FDDI configuration: FDDI servers and hubs. Hubs bridge to segmented >(48-72 nodes per channel) 10 mb ethernet. > >Fast Ethernet: 100 mb switch and NICs for servers. Switch has 10 mb >switched pipes out to the (10mb) hubs. Hub channels would have 48-72 >nodes per channel. > >I have always thought that FDDI would be better because of the higher >percentage of available bandwidth. I have spoken with others who say >that from a workstation standpoint, where the workstations are ethernet, >FDDI has more latency in the encapsulation used by the bridge from >ethernet to FDDI. With a DAS FDDI network you get more resilience as if one of the cables becomes faulty the nodes either side of it will warp their traffic around the ring. of course if the ring is all inside one building this isn't much of a gain. FDDI products are (possibly) more mature and will have software withmore features: eg Lanplex2500 ethernt/fddi switch started out as a plain switch, and now you can get IP/IPX/AppleTalk routing for it. the same has not YET happened for the LanPlex1000 devices. if you are going to build a large network with switches you must look carefully at the broadcast traffic, and decide if you really want it going everywhere. each time a workstation boots it will broadcast for the nearest fileserver/tree. you'll also need to lookat routing protocols and possibly move to less chatty ones such as OSPF and NLSP. for both FDDI and 100baseT you will get about the same latency from a switched-10Mbps port onto ring/100baseT as the switch will have to buffer the entire frame before it can be transmitted. going in the opposite direction again the latency should be about the same. I've not seen many ethernt/fddi devices offering cut-through, but then I'd suggest avoiding this on any moderately sized network (100+ machines) ------------------------------ Date: Tue, 01 Oct 1996 08:37:36 -0500 From: "Scott Etienne, C.N.E." To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet >What are your thoughts on FDDI vs. Fast enet? what about migration paths >to new(er) techololgies? does one of the technologies lend themselves to >a better migration path to the ATM, and is ATM going to be the real next >path or what about gig enet. Whatever we buy we have to live with forever. >They won't change for a long while. We are looking for the virtual >network so that students can log in anywhere and do what ever. > >Some tell me that FDDI is old, not scalable, and expensive. Others say >FDDI is like ATM and migrates nicely to ATM. There is an emerging technology called GB Ethernet. In it's current implementation it runs at 1.6 GB/Sec over fiber. I have seen advertisements from switch makers already offering it, and the prices are as low as fast ethernet. There is a review in the Sept. 16 issue of Infoworld of aGigaLabs Gigastar 2000 and smaller hub-size switches. If you are not comfortable with a new name, then maybe you can use this info as leverage with your vendor to find out who else has or is working on a similar solution. Otherwise, with switching, scalability is not really much of an issue anymore with FDDI or almost any other viable topology. Also, Fast Ethernet is only ten times faster than standard ethernet-plenty for even the fastest server, but maybe not enough for a multi-media backbone. Of course, you could look at a solution that has the promise of being upgradeable to something faster. --------- Date: Tue, 1 Oct 1996 11:41:57 EST From: "Robert L. Herron" To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet Unfortunately, too many people feel that ATM is still not ready for prime time. While ATM is "standardized," you must go with one equipment vendor for the least headaches. With FDDI, you can get a nice interbuilding backbone with built-in redundancy (assuming dual ring). However, FDDI has been around for several years and I have not heard of any improvements. But, you can have a total network span of up to 62 miles and internode span up to 25 miles. With Fast Ethernet, you get 100MB with backward compatibility. However, Fast Ethernet imposes more strigent distance and hop limitations. Regardless of which architecture you select, have fiber installed between the buildings. Running copper between buildings can create grounding problems and lightning will travel down copper quite nicely. Nothing ruins a good night sleep like being called into work to repair/recover from a lightning strike. Get quotes from some other VARs. When getting the quotes, list desired performance in the RFQ/RFP not the use of ATM, Fast Ethernet, or FDDI. Ultimately, the quote/proposal should explain why the VAR chose a specific architecture. Also, when design the network, consider the business functions that reside in each building. If Building B has a critical business function that would result is significant productivity lose, then Bldg B should have redundant connections to the network backbone. --------- Date: Wed, 2 Oct 96 09:18:43 -0700 From: Randy Grein To: "NetWare 4 list" Subject: Re: fddi vs. fast enet >Anyway. Anixter is showing a design of fast ethernet with limited atm to >a couple of buildings. Cabletron is fddi no ATM. > >Our campus is small 10 buildings within a city block. We currently have >no infrastructure in place. Hmm, that's interesting - Cabletron has ATM products as well, and some very nice ones, too. Could it be that your situation doesn't call for ATM? Please take the following with a grain of salt: ATM is HARDLY a finished standard. While the network adapters are interchangable, the switches are not - once you select a vendor, you're stuck for life. On the other hand, Fast ethernet has enjoyed strong growth in part because it does interoperate, is well understood and has good capabilities with full duplex mode; there is also yet another upgrade path on the horizon - gigabit ethernet. (Can you say switched full duplex gigabit ethernet to the desktop?) So much for raw throughput. The THEORETICAL advantages of ATM, namely time dependencies and quality of service are primarily for phone/video; unless you're trying to piggyback a lot of that junk onto the network it's really unnecessary. FDDI is rather expensive, as are most token passing schemes. It is very efficient over busy pipes and can be an excellent choice for a backbone consisting of numerous peers. FDDI can be used in a switched environment, but unless you're using it to connect several ethernet switches the benefits are limited. The upgrade path for any of these technologies is there - the interface is replaced at upgrade time so you're left with just the cable. The campus cabling WILL consist of fiber, so there really isn't much of an issue here. The long and short of the matter is that you're chosing between two good paths rather than two bad ones. I can say that Cabletron has a VERY interesting switch design; it was designed to be modular, extensible, and really lends itself to virtual networking. --------- Date: Tue, 01 Oct 1996 11:07:20 -0700 From: Darren Rogers To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet -Reply >>Anyway. Anixter is showing a design of fast ethernet with limited atm >>to a couple of buildings. Cabletron is fddi, no ATM. >> >>Our campus is small 10 buildings within a city block. We currently >>have no infrastructure in place. > >Hmm, that's interesting - Cabletron has ATM products as well, and >some very nice ones, too. Could it be that your situation doesn't call >for ATM? I'd have to second that question (if in fact one may second a question...). Many vendors out there try (very succesfully) to sell ATM because of all the press it's gotten, not based on client need. I personally would trust my Cabletron rep if he reccomended something other than ATM (and he did). I would get a couple other quotes (time permitting) and tehn sit down and have a long talk with your Cabletron rep. Ask him why he reccomended FDDI for your campus, and be sure that he knows you're not trying to change his mind. You may learn quite a bit. --------- Date: Tue, 01 Oct 1996 15:05:12 -0400 From: Sherri To: netw4-l@bgu.edu Subject: fddi vs. fast enet -Reply -Reply On 10/1/96, Darren Rogers wrote: >I'm gonna keep my mouth shut about the technology, but as far as >Cabletron is concerned, I'd hire them for almost anything. They are a >great company to work with, and their support is superb, and their >equipment is top drawer. I have to second this. I haven't had to address the FDDI vs. Fast ENet yet. But Cabletron has great tech support and the product has been very reliable. In the one unit that had one port die after only one year, they replaced it quickly with very little paperwork involved. I would recommend them anytime. --------- Date: Wed, 02 Oct 1996 10:53:53 +1100 From: Michael Cole To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet >Some tell me that FDDI is old, not scalable, and expensive. Others say FDDI >is like ATM and migrates nicely to ATM. > >Has anyone had experience with Bay vs. Cabletron as well? Pro's Con's. My two cents worth, also to be taken with a grain of salt; I have been in a Cabletron / Digital house before, with FDDI, MMAC8s' and their switching module. Had some problems switching Novell IPX Ethernet_802.3 into FDDI. Seemed to work OK with standard Ethernet_II and IPX Ethernet_802.2 frame types. Becareful of the Tron!! Bay has a reasonable 100BaseT story with enterprise and edge switches. As said before, fast ethernet is certainly more cost effective and you do not have any additional latency/packet loss in the translation from FDDI frames into ethernet frames. --------- Date: Tue, 01 Oct 1996 22:19:09 -0500 From: Darwin Collins To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet I like FDDI better than fast-Ethernet, because higher thruput and fault tolerence (dual ring). Since, FDDI is token-based you can almost get the full '100' out of the wire. Since, fast-Ethernet is still collision sensitive, you won't get more than about '30'. Also, myself, have found fast-Ethernet (100BaseT-x) to limited to very short distances. But, in anycase be sure to use Fiber if you are laying new cable. It simply a better long-term wiring plan. --------- Date: Wed, 2 Oct 1996 14:15:50 +0000 From: "Andrew Sear" To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet I was under the impression that once Fast Ethernet started using all 8 wires of the UTP (Cat 5) that you could get close to the '100'. I though it had something to do with the 'ACKS' coming back down a seperate pair. I also thought that there is a definite problem with 'Network Diameter' with fast ethernet ie. the 100metres from hub to workstation was no longer accurate. But as most people I have spoken to plan to use it with 10/100 switches so that the desktop gets 10mbps and the servers get 100mpbs the distance problems were a lot less significant. I thought you also had to have spot on Cat 5 cabling. I know longer have to follow this stuff as closely as I once did as our comms guys take care of it. As I have not seen anyone raise 100BaseVG-AnyLan, is it safe to assume that HP and IBM have backed a 'loser'? I thought their product sounded very impressive 12mths ago when I saw a demo, but it must have no market share or something. Please let me know if this product has any life left. HP were claiming that CSMA/CD was taking the worst feature of ethernet and keeping it with Fast ethernet and that their product was actually closer to ethernet than fast ethernet was. It would also run on cat 3 cabling and had a much wider network diameter. --------- Date: Wed, 2 Oct 96 22:38:53 -0700 From: Randy Grein To: "NetWare 4 list" Subject: Re: fddi vs. fast enet >I was under the impression that once Fast Ethernet started using all >8 wires of the UTP (Cat 5) that you could get close to the '100'. I >though it had something to do with the 'ACKS' coming back down a >seperate pair. I also thought that there is a definite problem with >'Network Diameter' with fast ethernet ie. the 100metres from hub to >workstation was no longer accurate. But as most people I have spoken >to plan to use it with 10/100 switches so that the desktop gets >10mbps and the servers get 100mpbs the distance problems were a lot >less significant. I thought you also had to have spot on Cat 5 >cabling. I know longer have to follow this stuff as closely as I >once did as our comms guys take care of it. 100Base-Tx uses 2 pair, not 4. There is a spec calling for cat 3 cable and 4 pair similar to 100VG, though. The funny thing about ethernet on tp is that there's a transmit pair, and a receive pair; they're connected only at the hub. Replace the hub with a switch, router or bridge and there's no collisions, and you can start contemplating full duplex. The network diameter problem (205 meters on UTP) is resolved as you've heard by switching. The cable must indeed be spot on, right down to patch cables and data jacks, but it's just not that hard to do. See, this is the beauty of ethernet - we have a transport that's cheap and easy with a statistical media access control mechanism. The cost savings produced by the cheap transport can be invested into switching technology, making it affordable, and we throw out the need for shared access. >As I have not seen anyone raise 100BaseVG-AnyLan, is it safe to assume >that HP and IBM have backed a 'loser'? I thought their product >sounded very impressive 12mths ago when I saw a demo, but it must >have no market share or something. Please let me know if this >product has any life left. HP were claiming that CSMA/CD was taking >the worst feature of ethernet and keeping it with Fast ethernet and >that their product was actually closer to ethernet than fast ethernet >was. It would also run on cat 3 cabling and had a much wider network >diameter. Oh, VG isn't dead - just not as popular. Problem is that it looks too much like Token-Ring no matter what HP says, and I think that scares people. The claim that it's closer to traditional ethernet than tx is funny, but the marketing wanks had a tough sell. Don't get me wrong - it's great technology, I just don't think it's destined for greatness. Most of the issues it solves are 5% ones - 5% of the people think they have these issues. --------- Date: Wed, 2 Oct 96 22:38:57 -0700 From: Randy Grein To: "NetWare 4 list" Subject: Re: fddi vs. fast enet >I like FDDI better than fast-Ethernet, because higher thruput and fault >tolerence (dual ring). Since, FDDI is token-based you can almost get >the full '100' out of the wire. Since, fast-Ethernet is still collision >sensitive, you won't get more than about '30'. Also, myself, have >found fast-Ethernet (100BaseT-x) to limited to very short distances. It's true that 100 baset-x is limited to 100 meters (205 meters total network diameter), but on fiber that's extended to several kilometers - I'd have to look the specs up to get the exact limit. As for performance I've seen too many reports showing over 70 mb/s to really believe your assertion of 30. Could it be that you're basing that on the traditional benchmark of 30% utilization for ethernet? If so, I'll happily contest that figure, having worked on many ethernet networks chugging along at 60% all day. It's quite true that FDDI will allow 90% + with good results resulting in a higher throughput, but it can't do full duplex - ethernet can. Of course this ONLY works point-to-point and is generally handled with switches, but that's the point. With a switching fabric in place and full duplex links you have an inexpensive technology that has a data rate of 200 mb/s. What's more, with full duplex collisions can't happen, and throughput can approach the theoretical maximum. Now if we throw gigabit ethernet into the picture, we have a technology that for sheer throughput blows ATM right out of the water. --------- Date: Wed, 02 Oct 1996 16:36:04 +1100 From: Michael Cole To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet >Oh, VG isn't dead - just not as popular. Problem is that it looks too >much like Token-Ring no matter what HP says, and I think that scares >people. The claim that it's closer to traditional ethernet than tx is >funny, but the marketing wanks had a tough sell. Don't get me wrong - >it's great technology, I just don't think it's destined for greatness. >Most of the issues it solves are 5% ones - 5% of the people think they >have these issues. Ditto, I don't think VG is destined for greatness, but it is interesting to note that although VG has little market penetration in the US or here in Australia, our European counterparts have taken it up with a little more enthusiasm (it runs over CAT3). There is also another fast ethernet story, 100BaseT4 (fast ethernet for CAT3 utp). 100BaseT4 utilises four pairs of voice or data grade cable and is a completely new standard as both 100BaseTx/Fx are based on ANSI FDDI technology. Personally I would upgrade your cabling to CAT5 and stick with 100BaseTx/Fx in a switched environment. --------- Date: Wed, 02 Oct 96 07:03:47 CDT From: John_Cochran@odp.tamu.edu To: netw4-l@bgu.edu Subject: Re[2]: fddi vs. fast enet I have to agree with Randy on this one. I just purchased a new Cabletron switch. It has 24 switched 10Mbps ports and 12 100Mbps (100base-TX) switched ports. This is going in my computer room to A) collapse my 10Mbps backbone to a flat switched 10Mbps and B) provide 100Mbps to most of my high-powered servers. Since this setup is capable of full duplex on both the 100Mbps and 10Mbps, I will be able to run my nightly backups at 200Mbps (Assuming that the server can push it to that level). Since there is no longer a collision domain on a switched port, I should be able to achieve this 'theoretical' throughput. In either case, it is much simpler to install. You dont have the cabling problems that you would have with fiber. And, since all of the machines are contained within a computer room, I dont have to worry about exceeding the length limitations on the 100baseTX. I was doing this same fast-ethernet to FDDI comparison several months ago. In our environment, it just didnt make sense to go FDDI. Now, if you have multiple buildings and need the redundancy of FDDI, then I would go with that. If you need more speed than standard FDDI, they even have switched FDDI which is much like switched ethernet! My recommendation would be to wait on the ATM... If you know you want to go to ATM, then install FDDI now to get your fiber plant in. But, wait until vendors figure out what they want to do with ATM. BTW, We are *very* pleased with Cabletron equipment. We use them exclusively for our network equipment and have for the past 5 years. Very stable hardware. --------- Date: Wed, 02 Oct 1996 09:02:58 -0500 From: "Scott Etienne, C.N.E." To: netw4-l@bgu.edu Subject: Re: Re[2]: fddi vs. fast enet >I have to agree with Randy on this one. I just purchased a new >Cabletron switch. It has 24 switched 10Mbps ports and 12 100Mbps >(100base-TX) switched ports. This is going in my computer room to A) >collapse my 10Mbps backbone to a flat switched 10Mbps and B) provide >100Mbps to most of my high-powered servers. XYLAN Rules. All of their Omni switches support mixed topology on the same backplane. Not only on the backbone, but all major topologies to the front panel. Supported topologies are: Ethernet, Fast Ethernet, Token Ring 4/16, FDDI fiber and copper, ATM fiber and copper. XYLan has also been instrumental in the development of VLANS. When everyone else had theirs in beta, all of XYLan's products were supporting them. You can have rules-based VLANs, port based VLANS, or MAC address based VLANS. There is also an available routing module, that makes any Omniswitch route packets in the 300k+ category- Cisco 7500 territory, I know-we have one of those too. Currently, we have ELEVEN XYLan Omni switches (two 9s and nine 5s), running on two 100mb FDDI rings. We have never had a problem. To say the least, no one on campus has any trouble getting around. The defense department is one of XYLan's biggest customers (VLANS are *VERY* secure). XYLan has an impressive array of top customers, and their competition is Cisco. Not bad for a garage-startup five years ago. I forgot to tell you about support. Every time we have done the smallest configuration changes to our switches, our XYLan SE's have come on site to help. They are always availble, are extremely intelligent and pleasant, and do nothing but help. We got caught with not enough flash ram on our switches to do a major OS upgrade. XYLan gave us the flash upgrade FREE. They even sent out an SE to do it for us (no charge)! Technically, we should have been charged, and I don't know why they didn't charge us. All I know, is that this company is a FAST riser because they have the best product out there and are really eager to please. And my boss loves'em! --------- Date: Wed, 02 Oct 1996 22:21:43 -0500 From: Darwin Collins To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet >It's true that 100 baset-x is limited to 100 meters (205 meters total >network diameter), but on fiber that's extended to several kilometers >- I'd have to look the specs up to get the exact limit. As for >performance I've seen too many reports showing over 70 mb/s to really >believe your assertion of 30. Could it be that you're basing that on the >traditional benchmark of 30% utilization for ethernet? If so, I'll Probably. But, I am told that the 30% utilization max still applies if you have more than 20 computers on the segment. The fewer computers that you have on a segment, the more likely, that you won't have the 'collision' storms. Perhaps, the 30% is not correct. Hmm, 100BaseT-X is for copper. (cat 5). I have never seen 100BaseT-X cards for Fiber. I learn something new all the time. >happily contest that figure, having worked on many ethernet networks >chugging along at 60% all day. It's quite true that FDDI will allow 90% + >with good results resulting in a higher throughput, but it can't do full >duplex - ethernet can. Of course this ONLY works point-to-point and is >generally handled with switches, but that's the point. With a switching >fabric in place and full duplex links you have an inexpensive technology >that has a data rate of 200 mb/s. What's more, with full duplex >collisions can't happen, and throughput can approach the theoretical >maximum. Now if we throw gigabit ethernet into the picture, we have a >technology that for sheer throughput blows ATM right out of the water. Hmm, so, it does depend on the environment. I can't give my 100mb/s boxes each a separate segment. But, I will remember your notes, for future sites. >Don't mean to roast you here, just wanted to set things straight. No problem. Keep it up. When we layed down our FDDI plant, 100BaseT was brand new, so we evaluated its 'then' limitations. I have once again, fell behind the technology. --------- Date: Thu, 3 Oct 96 23:47:54 -0700 From: Randy Grein To: "NetWare 4 list" Subject: Re: fddi vs. fast enet >Hmm, 100BaseT-X is for copper. (cat 5). I have never seen 100BaseT-X >cards for Fiber. I learn something new all the time. Technically you're right. I meant fast ethernet over fiber, but I think you understood that. --------- Date: Thu, 3 Oct 1996 08:46:48 -0600 From: "Mike Avery" To: netw4-l@bgu.edu Subject: Re: fddi vs. fast enet >>Uh, Darwin, it's true that 100 baset-x is limited to 100 meters (205 >>meters total network diameter), but on fiber that's extended to several >>kilometers - I'd have to look the specs up to get the exact limit. As for >>performance I've seen too many reports showing over 70 mb/s to really >>believe your assertion of 30. Could it be that you're basing that on the >>traditional benchmark of 30% utilization for ethernet? If so, I'll >Probably. But, I am told that the 30% utilization max still applies if >you have more than 20 computers on the segment. The fewer computers >that you have on a segment, the more likely, that you won't have the >'collision' storms. Perhaps, the 30% is not correct. I've gotten into a few discussions of this with technical staff at Thomas Conrad. Here's their spin on it. The 30% figure is from a truism that when your utilization exceeds 30% on ethernet doing more of whatever you did to get the utilization to that point will result in increased contention and overhead but not increased throughput. However, that truism dates back to Unix days, when Unix boxes were talking to one another as peers. Things have changed since then. Most of the time, for the people in this forum, we are dealing with NetWare servers and DOS or Windoze clients. And both the applications and the IPX/SPX protocol impose a meta-handshaking. As a result, utilization can get higher with little problems. As high as 60% in some cases. I've seen Ethernets at much higher utilizations, some happy, some not. With Ethernet, as with so many other things, your mileage may vary. Still.... Sniffer, LANAlyzer, and the Intel network monitor bundled with LANDesk by default all trigger alerts when the utilization exceeds 30% or 1000 packets per second. You can change these trigger levels, but those are the defaults. ------------------------------ Date: Tue, 21 Jan 1997 11:35:35 -0600 From: "Mike Avery" To: netw4-l@bgu.edu Subject: Re: ..need to vent against a hardware manufacture's blackmail >BEWARE... > >I assembled my lan 2.5 years ago with pieces provided by the school >board. > >This included Network Periferals FDDI cards and a Network Periferals >EIFO switching hub (made/distributed ? by Ungermann-Bass) for our >backbone. They had also bought may UB concentrators. > >One concentrator died.....*6* weeks for repair > >One FDDI card died...*ONLY* a 90 day waranty !!! My $ 70.00 D-link >10baseT cards have a lifetime warranty. > >My EIFO switch died...was told it only had a 90 day warranty. I >said for $ 8000.00 it better have a longer one. They got back to >me and revised their warranty on the switch to one year. I said we >would still like to have it repaired. They responded by telling me >THAT THEY WOULD NOT LOOK AT IT OR REPAIR IT UNTIL WE PAID AN UPFRONT >FEE OF $ 800.00 FOR A ONE YEAR SERVICE CONTRACT !!!!!!!!!. >This, in my mind, is akin to blackmail. I've used Network Peripherals cards in the past. Their FDDI cards used to have a lifetime warranty, and their service was excellent. They overnight advance shipped me replacements. However, since then their warranty period has dropped. And there is reason for it. Every FDDI card of theirs I have used has failed on me. Every one. And the word I've gotten is that their performance is not that good. Novell labs has found that the only FDDI card that can consistently deliver more than 60% of the nominally available bandwidth are from Madge. But.... here's a better option. Limp along with 10 base T, and as funds become available use 100baseT - I don't think that there is enough performance difference to justify the cost difference. --------- Date: Mon, 20 Jan 97 22:11:00 -0800 From: Randy Grein To: "NetWare 4 list" Subject: Re: ..need to vent against a hardware manufacture's blackmail >But.... here's a better option. Limp along with 10 base T, and as >funds become available use 100baseT - I don't think that there is >enough performance difference to justify the cost difference. Agreed, and if you get full duplex capable hardware it should outperform FDDI; at least the network will no longer be the bottleneck. And when it becomes the bottleneck again there's always load balancing, and gigabit ethernet on the near horizon. ------------------------------