[asterisk-users] Polycom phone cycles between UNREACHABLE and REACHABLE

[Rana Dutt]

I have a customer with a Polycom 501 phone behind a NAT. His phone is
connected to his Netgear router at home which in turn is connected to his
cable modem. The phone is set up to register with our remote Asterisk server
which is on a public, static IP address, with no NAT.

If we set qualify=yes, our Asterisk console shows his extension becoming
UNREACHABLE for a minute, then REACHABLE for a minute, then UNREACHABLE
again, in an endless cycle. If we try to call the phone while it is
UNREACHABLE, the phone never rings and the call goes straight to voice
mail. This is very annoying.

If we set qualify=no, then if we try to call the phone, the phone sometimes
does not ring at all, and we hear silence. The call eventually goes to voice
mail. This is equally annoying to the customer.

What is the solution to this problem? We have other customers with Polycom
phones behind NAT, and they don't have this problem. Will we have better
luck if we replace the Polycom with a Linksys 942 phone?

Here is some console output:

Jul 16 21:44:24 NOTICE[19981]: chan_sip.c:11364 sip_poke_noanswer: Peer
'280' is now UNREACHABLE!  Last qualify: 174
Jul 16 21:45:33 NOTICE[19981]: chan_sip.c:9697 handle_response_peerpoke:
Peer '280' is now REACHABLE! (3181ms / 5000ms)
Jul 16 21:47:37 NOTICE[19981]: chan_sip.c:11364 sip_poke_noanswer: Peer
'280' is now UNREACHABLE!  Last qualify: 175

Here is the way the phone is set up in sip.conf:

[280]
type=peer
username=280
secret=280
host=dynamic
dtmfmode=rfc2833
callerid="John" <280>
context=company_x
mailbox=280
nat=yes
canreinvite=no
qualify=5000

We are using Asterisk 1.2.5 with standard .conf files. We are not using
realtime or databases. Any help would be highly appreciated.

Rana Dutt
Softel Solutions
rdutt at softelinc.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20060716/40c54086/attachment.htm