>> Now, what is happening is this: the throughput is extremely "sluggish". If
>> I look at the 10BaseT Ethernet hub, what I observe is this:
>>
>> Now... has anybody of you *any* idea what is happening here? I've searched
>> the Kernel sources for something like "we have seen too many collisions,
>> let's back off and try again in a couple of seconds" or so, but couldn't
>> find anything.
>
>So maybe this is completely different, but then again ;),
We tried yesterday after reading Gert's announcement and the
problems that were reported with ppro/2.0.27/apache to see
if our server would hick-up.
We wrote a small program that will openup simultanious connections
to get a certain file from the www server.
The file in question is 1mb.
40 simultanious fetches resulted in a total bandwith of 920 kb/s coming
from the www server. starting up 200 sessions reported a total bandwith
of 750 kb/s. This was over a busy ethernet where other services were
active (appologis to all our quakeserver users who complaint about slow
response yesterday ;-)
I dont know what to make of it but it seems we have no real problem
in our setup. This particular server will start doing around 350.000 hits
per day with a growth of 10% per week. I will report back if we find
some special events.
Greetings,
Danny
-- Danny ter Haar | Cistron Internet Services | Unix & Internet dth@het.net | finger dth@cistron.nl for PGP-key | specialists ---------------------------+31-172-419445--------------------SP6------