Subject: Re: network throughput problems continue
To: Ron G. Minnich <rminnich@sarnoff.com>
From: Charles M. Hannum <mycroft@mit.edu>
List: port-alpha
Date: 09/04/1998 17:04:41
>>> so O(100K) packets, 0 collisions, 177 Ierrs. Things seem to work right,
>>> just very slowly.
>>
>> The large number of ierrs is indicative of late or undetected
>> collisions.
>
> is 117 ierrs out of 115046 Ipkts a large number? (serious question, I
> don't know).
Unless you're actively plugging or unplugging live machines, you
shouldn't see more than 1 or 2 per *week*. (And even in that case,
since you're using full-duplex mode, you must be using a switch, which
would eat any bad transmissions from another host.) 177 in a short
period indicates a serious network problem -- bad hardware, bad
cabling, etc. In this case I have to believe it's just a
configuration issue.