[netperf-talk] global question concerning Netperf test and SMP support

[Andrew Gallatin]

On 04/04/12 14:05, Rick Jones wrote:
> On 04/04/2012 10:58 AM, Andrew Gallatin wrote:
>>
>> Simon,
>>
>> A random few thoughts:
>>
>> Why do you use a 128 byte TCP message size? Unless there is some
>> particular use case you're interested in, this is going to stress
>> things like the client's system call interface, and not so much the
>> networking system. For 10GbE line rate, you're going to have to issue
>> more than 10 million syscalls / second. Also note the -- -M parameter
>> to adjust the *server*'s message size.
>
> Terminology nit - even if the netperf docs aren't good about it. :) The
> -m and -M options do not control message size, they control the size of
> the buffers posted in send and receive calls respectively. And in the
> case of -M, the receive calls can still complete before filling the buffer.

Indeed, sorry about that.  But if the OP is indeed testing some bizarre
use case of 128b messages, using the -- -M 128 flag will at least
cap the max receive to 128 bytes..

BTW, netperf -v 2 is very informative for this as it keeps track
of the average bytes/receive.

Drew