| Title: | FDDI - The Next Generation |
| Moderator: | NETCAD::STEFANI |
| Created: | Thu Apr 27 1989 |
| Last Modified: | Thu Jun 05 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 2259 |
| Total number of notes: | 8590 |
Hallo,
Imagine a 3rd vendor has got a TokenRing to Ethernet Bridge. This
Bridge implemented the Spanning Tree Algorithmus and works with the
Translation mechanism (instead of encapsulation).
Would it be possible that a Station connected to this Token Ring can
communicate with a Station connected to Ethernet(the Ethernet is
connected to the FDDI-Backbone with our DECbridge) via the FDDI Backbone?
Of course on both LANs they must use the same protocol (for example
TCP/IP).
I know the packet size on a Token Ring is higher than on a Ethernet.
But our Bridge does something like fragmentation?
I suppose you should use FDDI for such a application, but does it
really works?
Thanks Wiltrud
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 105.1 | �? | ZPOV01::HWCHOY | It must be Thursday. | Mon Aug 13 1990 21:38 | 7 |
Don't you mean TokenRing to FDDI? ---------------+
|
v
� Imagine a 3rd vendor has got a TokenRing to Ethernet Bridge. This
� Bridge implemented the Spanning Tree Algorithmus and works with the
| |||||
| 105.2 | CVG::PETTENGILL | mulp | Tue Aug 14 1990 21:56 | 20 | |
Once IBM backed down and agreed to `spanning tree' instead of `source routing' as the means of interconnecting LANs, it became possible to bridge between any arbitrary pairs of 802 compatible LANs and furthermore construct extended LANs which are composed of almost arbitrary topologies of different 802 LAN technologies. Do the bridges exist? I don't know beyond what DEC offers today, but I'm sure they will be produced by DEC and numerous other vendors as the market demands. Will the bridges operate like existing DEC bridges? Probably. DEC's profile is high enough and our products are functional enough to be the benchmark against which all other products will be measured. This means that future competing bridges can be expected to be 1) translating, not encapsulating and 2) will fragment IP packets when necessary. Has the problem of determining the minimum of all maximum packet sizes in a path been solved? No. Except for IP packets, the end nodes must either be set to the smallest maximum packet size on the LAN, ie., 1492 or the end nodes must know more about the LAN configuration than is apparent due to the transparent nature of the bridges. | |||||
| 105.3 | TRN<-->FDDI<-->ethernet really works? | FRAMBO::WILTRUD | Wed Aug 15 1990 02:47 | 7 | |
re: 105.1: Of course I'm talking of a Token Ring to FDDI-Bridge (sorry)
re: 105.2: So you suppose that communication Token
Ring<-->FDDI<--->Ethernet with Communication Protocoll TCP/IP should
work because our Bridge solves the packet Fragmentation for TCP/IP?
Is this the only problem you must care for if build such a Bridge?
| |||||
| 105.4 | KONING::KONING | NI1D @FN42eq | Wed Aug 15 1990 18:39 | 13 | |
There are two places in that picture where packet size issues occur: (1) FDDI to Ethernet -- we handle that for TCP/IP; (2) 802.5 to FDDI -- since 802.5 has a larger max packet size than FDDI. The same solution would work, of course, and whoever builds the 802.5 to FDDI bridge will presumably apply it. For 802.5 there are various other problems that an 802.5 to FDDI (or 802.5 to anything else) bridge has to deal with, such as incorrect address encoding in ARP packets in most 802.5 implementations, "functional" addresses used rather than real multicast, and so on. FDDI to Ethernet is pretty simply by comparison! paul | |||||