Subject: Re: Binding more than one IP to a NIC
To: Justin Newcomer <liquidice5@gmail.com>
From: Andy Ruhl <acruhl@gmail.com>
List: port-cobalt
Date: 11/07/2004 06:48:13
On Sun, 7 Nov 2004 02:53:53 -0500, Justin Newcomer <liquidice5@gmail.com> wrote:
>                     /0   /1   /2   /3   /4   /5   /6   /7   /8   /9   /10
>      Load Average   ||||||||||||||||
> 
>               /0   /10  /20  /30  /40  /50  /60  /70  /80  /90  /100
>     cpu  user|
>          nice|
>        system|XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
>     interrupt|
>          idle|
> 
>               /0   /10  /20  /30  /40  /50  /60  /70  /80  /90  /100
>     wd0  KBps|XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX2903.0
>           tps|XXXXXXXXXXXXXXXXXXXXXXX
> 
> I dont understand all that hibbidy bibibdy mumbo jumbo with ttcp, but
> i copied a 800 meg file to my qube 2, 1.6.1 netBSD with samba, then
> copied it back with apache.  the numbers were all pretty much 3MB/s +
> - 500ish KB
> 
> I didnt do anything fancy to my qube, cause I dont think i would know
> how to tweak the drivers lol, so if this isnt as fast as you would
> like, im curious to see if we should be expecting more than this? it
> seems good enough to me

I got:

ttcp-t: 733945856 bytes in 563.28 real seconds = 1272.45 KB/sec +++

Christian got:

ttcp-t: 858750976 bytes in 711.51 real seconds = 1178.66 KB/sec +++

Check the rates. They are 100 KB/s off or so, but close. Not so good though.

What I did was cat a file from a NetBSD/i386 machine (1 ghz Athlon,
100baseT Intel Pro card) to the Qube2. This same file going to my
FreeBSD 4.10 server (dual 366mhz Celerons, 256 megs of memory,
100baseT Intel Pro card):

ttcp-r: 733945856 bytes in 92.65 real seconds = 7736.41 KB/sec +++

Quite a bit quicker. This is on the same hub. Today was slower than
yesterday probably because I'm doing more stuff on the source machine.
But still way faster than the Qube.

Apparently the Qube with it's native OS (Linux based) doesn't have this problem.

Andy