| 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 |
Does anyone have a clear understanding of Digitals implementation of
the ANSI FDDI standards for connection rules?
The question I have is;
What is the probability that the FDDI connection rules for DAS attachment
to M ports can be violated?
From what I have deducted, if the rules are violated, both A and B
ports could be activated simultaneously resulting in the removal of
downstream stations from the ring. If this is possible then there could
be serious problems with the FDDI reliability.
Another question:
If a station's port is stuck in a given line state, the connected
concentrator's port can become stuck. The result will be that the
connection will never initialize. I assume that the concentrator's
management function will dedect the failure and remove the M port from
the data path. Correct?
Last question:
Station ManagemenT. Anyone know which version of SMT Digital complies to?
Phil.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1245.1 | connections rules | QUIVER::PARISEAU | Luc Pariseau | Wed Feb 23 1994 09:17 | 13 |
I'm not sure what you mean. Are you asking "What happens if an SMT implementation violates the Dual Homing rules?". This will result in a third ring being created (bad thing.) What is the probability?? 0 on a good implementation, High on a bad one... The M port will not insert a station onto the data path until the connection completes successfully. So yes, correct. If the product has DEC's latest version, then it implements SMT 7.3. Luc | |||||
| 1245.2 | KONING::KONING | Paul Koning, B-16504 | Wed Feb 23 1994 11:37 | 17 | |
The connection rules are enforced in the station firmware. I've seen no evidence of bugs in that area, but of course those algorithms can have bugs in them just like any other product, software or hardware. The algorithm is simple and deterministic, so you can test it and convince yourself it's right. As for stuck line states, selftest will catch that if it occurs prior to selftest (or if you reinitialize the station). If it happens later, but before connection, the connection procedure will fail. If it happens after connection, this will cause a PC Trace to be done, which will cause a selftest, which will cause the defective port to be shut down. So in all cases the fault will be caught. In all but the last, it will have NO effect on the rest of the network (other than the obvious point that the defective station is off the net); in the last case, the ring will be disrupted for the time it takes PC Trace to be done, which is perhaps 30 seconds. paul | |||||