<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<blockquote>If Asterisk starts and no DNS is available, outbound
registrations will fail indefinitely. This patch copies the
address from the sip_registry struct, which will be updated, to
the peer->addr when necessary.
If dnsmgr is enabled, the registration fails without the patch
because even though the address on the registry is updated via
dnsmgr, the address is just copied on the first try. Since we use
ast_sockaddr_copy, dnsmgr can't update the address that is copied
to the sip_pvt or peers.<br>
<br>
</blockquote>
<pre style="margin: 0; padding: 0; white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">Just thought I would throw in a comment regarding 1.6.2 in case you guys weren't aware. If DNS is enabled, and working, and internet connectivity is lost, the DNS lookups will timeout (30 second delay or more) and asterisk will go not responding during that time. In my case, there are POTS lines that are used for failover when the SIP lines go down and the DNS lookup attempts (for the SIP trunk peer) make it so you can't even answer the POTS line (it rings, but you pick up the phone and it won't bridge the call) because Asterisk is so bogged down. I have changed my SIP trunk peer to an IP address to work around the problem, but if they change their IP my service will go down.
This may have already been addressed in newer versions, or maybe there is a way to shorten the DNS timeout time to a couple seconds.
</pre>
</body>
</html>