Hi Alex, You need to set the Nat Keep-Alive to yes in the Ext config for the SPA 303's.
Cheers, Doug -----Original Message----- From: Alex Robar [mailto:[email protected]] Sent: November-02-12 9:49 AM To: [email protected] Subject: [on-asterisk] Local SIP Traffic Disrupted with Bell Sagemcom F@st 2864 Fibe Modem/Router Wondering if anyone has experienced issues with the Bell Sagemcom F@st 2864 modem/router that's being shipped with newer Fibe services? Ran into an issue yesterday that I can't wrap my head around. Configuration looked like this: Bell Modem -> Switch -> Phones & Asterisk DHCP leases came from the Bell modem. Phones were configured to use Asterisk for primary DNS, Bell Modem as their secondary DNS. Phones are Cisco SPA 303s. On a SIP reload, all phones would register properly to the local Asterisk instance. Calls could be made between extensions. After about 30 seconds, all peers showed a status of "UNREACHABLE". Obviously at that point, calls no longer worked. Take the Bell Modem out of the equation and put in something else to hand out leases, everything works just fine. The configuration is identical on the other router (same DNS forwarders, IP scheme, etc.). Traffic captures on Asterisk made it appear as though traffic was just disappearing. Initially we'd see the SIP registrations, options packets went back and forth, etc. After the 30 second period, we would send challenges in response to re-registrations, but nothing came back. However, Asterisk can still ping the phones at this point. Turned on remote syslogging for the phones and rebooted them a couple of times. The phones simply showed "SIP:RegFailed;Retry in 30s". The logs show that the IP the phones are connecting to is the correct, internal IP address of the Asterisk server. So from what I can see, it looks like Asterisk says it's sending the challenge, and the phones say they're sending responses. Both sides, despite being able to route packets to one another, state that the other side did not respond. As with most ISP gateways, the fix was to put a different router in front of the modem and put the modem into bridge mode - but I'm stumped as to what could have caused LOCAL traffic to get dropped when the Sagemcom gateway is in place. Why is traffic that should only ever have to cross the switch, and not the gateway, getting redirected or eaten by the gateway? What mechanism is the gateway using to do this? Any thoughts appreciated! Thanks, Alex -- Alex Robar [email protected] ... -- BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach the Toronto Phone & Data Anti-Spam if this mail (ID 01IiBNdbw) is spam: Spam: http://antispam1.torontophoneanddata.com/canit/b.php?i=01IiBNdbw&m=cef92234bfea&c=s Not spam: http://antispam1.torontophoneanddata.com/canit/b.php?i=01IiBNdbw&m=cef92234bfea&c=n Forget vote: http://antispam1.torontophoneanddata.com/canit/b.php?i=01IiBNdbw&m=cef92234bfea&c=f ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS ... --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
