<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>cpu affinity? If you have network card with many tx/rx threads
(modern PCI-E card can use MSI-X and 'software irq'), you can bind
different card threads/irqs to cores and ntpd process to other
core. On BSD we use cpuset to spread and bing threads to cores. <br>
</p>
<p>On Linux see script set_irq_affinity.sh<strong> </strong>from
Intel drivers (<a
href="https://gist.github.com/SaveTheRbtz/8875474">https://gist.github.com/SaveTheRbtz/8875474)
</a>and others in in drivers <a
href="https://sourceforge.net/projects/e1000/files/">https://sourceforge.net/projects/e1000/files/</a></p>
<p>Also you can google articles like 'linux router performance', for
example <a href="https://github.com/strizhechenko/netutils-linux">https://github.com/strizhechenko/netutils-linux</a>
(maybe also rss-ladder tools can help ) or <a
href="https://blog.packagecloud.io/eng/2016/06/22/monitoring-tuning-linux-networking-stack-receiving-data/">https://blog.packagecloud.io/eng/2016/06/22/monitoring-tuning-linux-networking-stack-receiving-data/
<br>
</a></p>
<p>Network stack tuning not simple. Performance is need good NIC
multithread chip and good driver. As I know, Intel NIC chipsets
and drivers really the best here.<br>
</p>
--<br>
Mike Yurlov<br>
<p><br>
</p>
<div class="moz-cite-prefix">13.01.2020 11:54, Hal Murray пишет:<br>
</div>
<blockquote type="cite"
cite="mid:20200113085416.537CE40605C@ip-64-139-1-69.sjc.megapath.net">
<pre class="moz-quote-pre" wrap="">Thanks.
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">and without 'limited' on ~5kpps I have 8-10% CPU regardless minitoring
enabled/disabled. About 1% on 1000pps.
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">
Is that within reason or worth investigating? 1% times 5 should be 5% rather
than 8-10% but there may not be enough significant digits in any of the
numbers.
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">For those who want to process hundreds of thousands of requests per second
(like 'national standard' servers) you can use multithreading and multiply
power of server.
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">
The current code isn't setup for threads. I think with a bit of work, we
could get multiple threads on the server side.
On an Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz
I can get 330K packets per second.
258K with AES CMAC.
I don't have NTS numbers yet.
</pre>
</blockquote>
</body>
</html>