Re: /netbsd: ser0: silo overflow

["Georges Heinesch" <geohei-ml%geohei.lu@localhost>]

Quoting Ignatios Souvatzis (04-May-01 17:10:55):

> On Fri, May 04, 2001 at 08:39:39AM +0100, Georges Heinesch wrote:
>> 
>> Name  Mtu   Network       Address               Ibytes     Obytes
>> ppp0* 1500  <Link>                             2778797     465895
>> ppp0* 1500  fe80::        fe80::260:30ff:fe    2778797     465895
>> 
>> dmesg | grep silo | wc
>>     27  x   x

> So that's one lost packet) per 100000 bytes transferred? And you're 
> complaining?

Till now, I thought that when RTS/CTS is used, 100% throughput is
guaranteed, but if you show it as a "good yield" in terms of "losses
over general throughput", then I agree that it's a satisfying result!

I was quite surprised to see this error at all.

As far as I understood now, "the silo error can appear from time to
time without any reason to worry."

-- 
Cu  Georges Heinesch, Luxembourg
    geohei%geohei.lu@localhost
    http://www.geohei.lu
    PGP RSA & DH/DSS public key on request and on public servers

... N1, TO/GA, 80 KTS, V1, ROTATE, Postive climb, Gear up ...