how-to block ads
|
tempnexus
Premium
join:1999-08-11
Boston, MA
| Network error handling. OK I got a question this seems like a standard practice in networking but I cant find any reference to back up my assumption. So if anyone could help me find a reference that teaches:
"Once the Error code (or errors)are received over the network, the Ethernet link (card) waits a specific period of time, and if the errors are still present after the specific amount of time, it disables the ethernet port (in transmission direction) (or shuts it down or redirect it etc)" next "Enables the ethernet card/port/transmission once the error codes (Errors) no longer are being received, but before it does that the card waits a specific amount of time before re-enbling the tranmission just to make sure that the error free state is not transient).
Thanks a bunch, want to win an argument by telling my friend that what he is saying is soo obvious that it has been done before...except I can't find any reference to support my statement.
Mods: If this is inapropriate forum them could you please move it to one that fits this topic the best. Thank you so much. | |
|  dave
Premium,MVM
join:2000-05-04
not in ohio | Re: Network error handling. At present, it's too vague to comment on.
E.g., what does "error code ... received" mean?
What protocol are we talking about? | |
| |
tempnexus
Premium
join:1999-08-11
Boston, MA | Re: Network error handling. error code = erronous packets, missed packets, crc failed etc.
TCP or Ethernet over Sonet, Sonet, etc. | |
| dave
Premium,MVM
join:2000-05-04
not in ohio
 ·Verizon Online DSL
 ·Verizon FIOS
| Oh, nothing that was actually received, then - you mean that the adapter reported an error?
So the theory is that the NIC stops transmitting on error condition, and restarts transmitting when the error condition has cleared. And the allegedly novel idea is that the NIC should 'wait a bit longer' just to make sure the error condition has really cleared?
You're right, it doesn't sound too earth-shattering.
I think the crux of the matter hinges on how the state 'error condition has cleared' is defined? You may find that the definition includes some time consideration -- e.g., 'such and such a signal has remained clear for at least X microsecs'.
I have no reference for you, though. | |
| | | |
|
