m=E5ndag 03 november 2003 14:02 skrev du:
> Hi,
> I had a similar behavior some days before. That damn gereric rtk NIC
> felt more like a 56kp/s modem, than a fast-ethernet card. After fixing it
> to 100Mb/s half-duplex mode, problems went away. Maybe this can help
> you too.
>
> Whith "netstat -s" on a DOS-Box you can take a look on built-in packet
> statisics. Somewhere odd error counts?
>
> And there is this (in)famous "MTU to big and path MTU detection failed"
> issue, which is always capable to mess things up. How to work around this
> with MMS clamping on the gateway, is discribed in the pppoe(4) man page.
> Or you set it down by hand on you windows host with tools like Dr. TCP.
>
> Ok, this does not explain why SSH'ing to the NetBSD host (It is on your
> LAN isn't it?) also get's messed up but I ran out of ideas now ;-)
>
> HTH
>
> -Alex

Thanks for the reply.
Yes, the NetBSD host is the gateway between my network and the outside and =
is=20
a part of my network. I've tried to repeat last nights situation and today=
=20
the XP box seems to be in a good mood. Everything works smoothly, including=
=20
the SSH part. That makes it difficult(at least for me) to find out what the=
=20
problem was with last nights bad performance, since I havn't changed anythi=
ng=20
on any of the machines involved...
I've been presented the possibility that it could be a worm but that doesn'=
t=20
seem to be the case from what I can see, at least not today.