And the Cretin-Of-The-Day award goes to .... &#39;me&#39;. <div><br></div><div>Having re-looked at the Wireshark trace I realised I hadn&#39;t opened up enough ports in the FreeSWITCH firewall to let the RTP from the Polycom phone through - so it was unable to learn what the real public IP address and port number.</div>
<div><br></div><div>This is what I get for breaking my own rules about working too late!</div><div><br></div><div>Forget I asked....move along....nothing to see here.<br><br><div class="gmail_quote">On Fri, Jan 21, 2011 at 10:44 PM, Mike <span dir="ltr">&lt;<a href="mailto:motosota@gmail.com">motosota@gmail.com</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Scenario. Polycom phone, private IP address, behind Cisco router doing NAT overload to a public IP address (SIP &#39;fixup&#39; disabled, so the Cisco isn&#39;t mangling anything). <div>
<br></div><div>RTP stream from FreeSWITCH is sent to the private, not public address of the phone.</div>
<div><br></div><div>I&#39;ve got this working on 1.0.6  but I&#39;ve tried this with FreeSWITCH Version 1.0.head (git-7070061 2011-01-20 13-52-00 -0600) and it doesn&#39;t seem to work for me.<div><br></div><div>Here <a href="http://pastebin.freeswitch.org/15109" target="_blank">http://pastebin.freeswitch.org/15109</a> is my sip profile.</div>

<div><br></div><div>Here <a href="http://pastebin.freeswitch.org/15110" target="_blank">http://pastebin.freeswitch.org/15110</a> is a FreeSWITCH log for a call made from one of the Polycoms to voicemail with SIP tracing enabled.</div>
<div>
<br></div><div>And here <a href="http://pastebin.freeswitch.org/15111" target="_blank">http://pastebin.freeswitch.org/15111</a> is the wireshark trace showing the media stream being sent by FreeSWITCH to the wrong port.</div>
</div><div><br>
</div><div>I may be (hopefully) missing something very obvious here. </div><div><br></div><font color="#888888"><div>Mike</div>
</font></blockquote></div><br></div>