On Thu, 21 Jun 2001, Bill Woods wrote:
> I have need to setup a small lab designed to saturate a pipe. The
> plan is to use Web Polygraph like shown below
>
> <<...OLE_Obj...>>
Sorry, my mailer cannot display your embedded(?) picture. What you see
above is what I get.
I assume you meant something like this:
clt1 --------\ /-------- srv1
clt2 ---------\ /--------- srv2
switch1 ===pipe=== switch2
clt3 ---------/ \--------- srv3
clt4 --------/ \-------- srv4
> The pipe between the switches is a 100 MBPS Full Duplex LAN
> connection (This is the connection I need to saturate). The 4
> systems on the left are going to be servers, while the 4 on the
> left are clients. Is this something Web Polygraph would be
> suitable for ?
Sure. However, keep in mind that Polygraph is best at generating
realistic Web traffic, not saturating pipes with stupid packets.
Whether Polygraph is the right tool for you would depend on why
you need to saturate that pipe. In other words, it would depend
on what you want to look at after/while the pipe is saturated.
> Would Web Polygraph be able to generate enough web traffic
> simulation to accomplish this? All Clients and servers are
> connected to the switches by 100 MBPS Full Duplex NIC Cards.
Yes. You can saturate a 100MBit link with one or two client-server
pairs, depending on your CPU speed. Note that Web traffic is very
asymmetric so in a typical setup you will not see 100Mbit streams
going in both directions.
> The plan also calls for the servers and clients to be running
> either Linux or FreeBSD
Both should work. We use, recommend, and, hence, may be able to help
with FreeBSD.
Good luck,
Alex.
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:19 MDT