<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>my perftest suite<br>
</p>
call generator<br>
sipp, but creating sipp scenario is not easy. i'm using more user
friendly(but its for win) - <a class="moz-txt-link-freetext" href="http://startrinity.com">http://startrinity.com</a> (REST API
available)<br>
<br>
device emulation<br>
using asterisk as SIP client in docker - 30 SIP endpoints per
instance<br>
<br>
reports<br>
pbx cpu/load/.. - grafana<br>
call details - export from startrinity<br>
<br>
my plan for automation is node.js app connected via ARI to Asterisk
and via REST API to startrinity. endpoints configuration via ARI
Push to PJSIP realtime<br>
<br>
<br>
<br>
<div class="moz-cite-prefix">Dne 20/09/2017 v 13:49 Olivier
napsal(a):<br>
</div>
<blockquote type="cite"
cite="mid:CAPeT9jjbgAyunqwtkvnQFXMeqoX1885O8h8BSY-fZN70rP83iA@mail.gmail.com">
<div dir="ltr">
<div>
<div>
<div>
<div>Hello,<br>
<br>
</div>
I am currently tasked on how to load test both signal and
media from a couple of Asterisk machines which are doing
corporate SIP trunking (no phone endpoint).<br>
<br>
</div>
<div>If that matters, ecah machine will host debian Stretch,
Asterisk 13 with either classic SIP or PJSIP.<br>
<br>
</div>
For instance, I can generate from a given source machine to
a destination machine, 1000 calls passing an Asterisk
instance under test.<br>
<br>
</div>
This under test Asterisk instance generate CDRs in which a
testing program can seach for successful calls (reading
disposition and/or RTCP stats in userfield).<br>
<br>
</div>
<div>Most probably, this under test Asterisk instance will also
log SIP capture to a remote Homer server, using Capagent and
Homer itself.<br>
</div>
<div>A testing program can also search this Homer/Capture
database to evaluate "testing exit code".<br>
<br>
</div>
<div>My question are:<br>
<br>
</div>
<div>1. Which (preferably available on Debian) tool(s) would you
use to assert a single captured call, recorded on purpose by
the system under test, has met call quality requirements ?
(This one-call tool is needed to calibrate next tool).<br>
<br>
</div>
<div>2. Which tool(s) would you use to do the same, on whole
testing campaign generating 1000 or 2000 simultaneous calls ?<br>
<br>
</div>
<div>3. How can you automate such tests ?<br>
<br>
</div>
<div>Cheers<br>
</div>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
</blockquote>
<br>
</body>
</html>