<div dir="ltr">Hi Alex-<div> I wonder if this would help you gather some data:</div><div><br></div><div><a href="http://oss.oetiker.ch/smokeping/">http://oss.oetiker.ch/smokeping/</a><br></div><div><br></div><div>(written by Tobi of RRdtool fame)</div><div><br></div><div> -- dNb</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Aug 24, 2015 at 1:24 PM, Alex Aminoff <span dir="ltr"><<a href="mailto:alex@basespace.net" target="_blank">alex@basespace.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Our situation is that our network becomes slow at random times. We have looked at obvious things to look at, like the IO utilization and CPU and memory on the file server and we have attempted to look for error rates on switches, with no obvious useful result.<br>
<br>
What I would like to do is have one or more packet capture files from when the network is running smoothly, and then one or more samples from when it is not, and some sort of software to compare them statistically. Maybe it would produce a report of the types of packets whose frequency changed the most.<br>
<br>
Surely something like this must exist? It is conceptually (mathematically?) similar to bayesian spam detection, I would think.<br>
<br>
- Alex<br>
<br>
_______________________________________________<br>
bblisa mailing list<br>
<a href="mailto:bblisa@bblisa.org" target="_blank">bblisa@bblisa.org</a><br>
<a href="http://www.bblisa.org/mailman/listinfo/bblisa" rel="noreferrer" target="_blank">http://www.bblisa.org/mailman/listinfo/bblisa</a><br>
</blockquote></div><br></div>