<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p>Hi,</p>
<p><br>
</p>
<p>I took the liberty of setting up a couple of measurements using 6
of the probes you used, but with a starting hop of '1' so I could
see the paths taken.</p>
<p><br>
</p>
<p>TCP: <a class="moz-txt-link-freetext" href="https://atlas.ripe.net/measurements/28919237/#probes">https://atlas.ripe.net/measurements/28919237/#probes</a></p>
<p>ICMP: <a class="moz-txt-link-freetext" href="https://atlas.ripe.net/measurements/28919338/#probes">https://atlas.ripe.net/measurements/28919338/#probes</a></p>
<p><br>
</p>
<p>None of the probes receive a response over ICMP (this is not
uncommon), and five of the probes receive a response over TCP. Be
aware that the UI will report the RTT for the last responsive hop
in either case, *whether it's the target or not*. The actual
result data reveals a more.<br>
</p>
<br>
<p>In the case of probe 17832, it doesn't reach the target in either
case above; the low RTT is the last responsive hop very close to
the probe.<br>
</p>
<p><br>
</p>
<p>17875 and 27091 take more reasonable paths over ICMP to the
target, but over TCP appear to receive a response from the target
at the second hop. In terms of what's handling that traffic, I
don't know, but with latencies so low it's not very far upstream
at all.<br>
</p>
<p><br>
</p>
<p>S.</p>
<p><br>
</p>
<p><br>
</p>
<p><br>
</p>
<div class="moz-cite-prefix">On 27/01/2021 14:41, Hank Nussbacher
wrote:<br>
</div>
<blockquote type="cite"
cite="mid:133db351-9060-1893-223b-6ea2ac2478f8@interall.co.il">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<style id="bidiui-paragraph-margins" type="text/css">body p { margin-bottom: 0cm; margin-top: 0pt; }</style>
<p>I found this presentation:<br>
<a class="moz-txt-link-freetext"
href="https://www.ripe.net/support/training/ripe-ncc-educa/presentations/measuring-reachability-of-your-web-server-using-ripe-atlas.pdf"
moz-do-not-send="true">https://www.ripe.net/support/training/ripe-ncc-educa/presentations/measuring-reachability-of-your-web-server-using-ripe-atlas.pdf</a><br>
<br>
I ran a test from 40 Israeli probes to see if there are any
issues accessing zoom.us:<br>
<a class="moz-txt-link-freetext"
href="https://atlas.ripe.net/measurements/28912347/#probes"
moz-do-not-send="true">https://atlas.ripe.net/measurements/28912347/#probes</a>
[1st run]<br>
<a class="moz-txt-link-freetext"
href="https://atlas.ripe.net/measurements/28912916/#probes"
moz-do-not-send="true">https://atlas.ripe.net/measurements/28912916/#probes</a>
[2nd run]</p>
<p><br>
</p>
<p>It shows all in Israel getting to Zoom.us with an average of
150ms. <br>
But 3 sites – show under 1ms response time – as if they run
some local proxy. I checked with them and they said they have
nothing local.</p>
<p>So how come probe 17832 gets .869ms and probe 17875 gets .610ms
and probe 27091 gets .718ms to zoom.us?</p>
<p><br>
</p>
<p>Thanks,</p>
<p>Hank<br>
</p>
</blockquote>
</body>
</html>