[netperf-talk] Testing simultaneus connections

[Gerd v. Egidy]

Hi Rick,

thanks for your quick response.

> > I want to measure the number of new tcp connections a firewall can handle
> > within a given time.
> 
> Perhaps an overly-subtle nit, but that is a connection rate, not the
> simultaneous connections from the subject.

oh yeah. I'll probably have to measure both.

I guess that initiating a new connection is much more expensive than just 
sending data over an existing one or even having an idle connection. So new 
connection rate comes first. Then I'll have to make up some data on how long a 
connection is open and how much data is transferred to be able to test for a 
number of concurrent connections.

On the other hand a big number of idle connections will make ram exhaustion 
and maybe hash collisions a problem. 

So you are right, those are two different tests.
 
> > Does netperf4 already support something like this? Or would it be
> > possible to add it with reasonable effort?
> 
> Reasonable is relative.  Netperf4 hasn't had much of anything done to it
> in a very long time and may suffer from non-trivial bit-rot.

Ok, thanks for the hint. So I'll try netperf2 first.
 
> That said, if the question is TCP connections through a firewall, do the
> TCP connections actually have to be carrying HTTP?

No, that was just an example. A plain TCP connection is enough.

> If not, the standard
> TCP_CRR test (not sure if there is one in netperf4 already, but it is in
> netperf2) could be used, along with the mechanisms one can use to get
> reasonable aggregate results from netperf2 (which is one reason netperf4
> hasn't had as much attention...) described in
> http://www.netperf.org/svn/netperf2/trunk/doc/netperf.html#Using-Netperf-to
> -Measure-Aggregate-Performance
> 
> These days I find myself using the technique described in Section 7.3 -
> http://www.netperf.org/svn/netperf2/trunk/doc/netperf.html#Using-_002d_002d
> enable_002ddemo combined with rrdtool - I use a "demo interval" of 1 second
> and a one second step in rrd.

Ok, tomorrow I'll play a bit with these options. I think I will have to come 
up with some source ip selection/distribution scheme. But a simple static 
distribution should be doable in a shell script.
 
> If you need actual HTTP traffic, you might consider httperf
> http://www.hpl.hp.com/research/linux/httperf/  Or you could in theory
> script something around the likes of curl or wget.

Thanks for the pointers. Currently I use siege http://www.joedog.org/siege-
home/ for the tests that really need http.

Kind regards,

Gerd