<div><div dir="auto">Hi Junior</div></div><div dir="auto"><br></div><div dir="auto">2300 pps is about the performance limit of a single NFD instance.</div><div dir="auto">See <a href="https://redmine.named-data.net/issues/3564#note-24">https://redmine.named-data.net/issues/3564#note-24</a> item 2.</div><div dir="auto"><br></div><div dir="auto">You can make a dumbell topology of 9 machines: 4 consumers, 1 bottleneck forwarder, and 4 producers.</div><div dir="auto">Each consumer node can send 2300 pps. When it gets to the bottleneck forwarder, it can probably process more traffic: 8600 pps.</div><div dir="auto"><br></div><div dir="auto">Yours, Junxiao</div><div><br><div class="gmail_quote"><div dir="ltr">On Tue, Sep 11, 2018 at 08:47 Junior DONGO <<a href="mailto:junior.dongo@lacl.fr">junior.dongo@lacl.fr</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<p>Hello,</p>
<p>How can I perform NFD stress test? <br>
</p>
<p>I am using <i>ndnping -i 1 /example/testApp</i> to generate
interests, but not able to have many packets per second. Running
for example 5 processes in parallel, the global number of packets
per second to be transmitted should be around 5000, but only
around 2300 packets per second are transmitted from the interface
of the running node. <br>
</p>
<p>Any idea?<br>
</p>
<p>Regards,</p>
<p>Junior</p></div></blockquote></div></div>