We had major memory leak problem with version 13.0. The version 13.2 fix this big problem. Good news! But after upgrading from Asterisk 13.0 to 13.2, many contacts was reported “unknown nan” with the cli command “pjsip show contacts”. After rebooting endpoints (Polycom), they are reported OK but not all. Here what we see continously in the log for these contacts:
[2015-02-20 13:32:42] ERROR[6953]: res_pjsip/pjsip_options.c:315 qualify_contact: Unable to apply outbound proxy on request to qualify contact sip:280@X.X.X.X:5260
But all is working fine, users can receive and make calls without problem.
We don’t use NAT and we don’t have option “outbound_proxy” set. I don’t know why pjsip driver push this option. The port 5060 is use. I have try to delete and recreate one pjsip extension, rebooting the server but the problem still there.
I have look in Asterisk’s AstDB and I see the problem:
Here for bad contact:
/registrar/contact/193;@sip:193@X.X.X.X:5360: {“outbound_proxy”:“asterisk”,“uri”:“sip:193@X.X.X.X:5360”,“user_agent”:“Linksys/SPA8000-5.1.9(a)”,“expiration_time”:“1424460007”,“qualify_frequency”:“60”,“path”:“asterisk”}
Here for good contact:
/registrar/contact/197;@sip:197@X.X.X.X:5060: {“outbound_proxy”:"",“uri”:“sip:197@X.X.X.X:5060”,“user_agent”:“PolycomSoundPointIP-SPIP_550-UA/3.3.5.0247”,“expiration_time”:“1424460266”,“qualify_frequency”:“60”,“path”:""}
In the first one there is a outbound_proxy setting but I know where it comes from.
Have you some clue or advice?
pjsip.transport.conf
[0.0.0.0-udp]
type=transport
protocol=udp
bind=0.0.0.0:5060
local_net=X.X.X.0/24
pjsip.endpoint.conf
[280]
type=endpoint
aors=280
auth=280-auth
allow=ulaw,alaw,gsm,g726
context=from-internal
callerid=device <280>
dtmf_mode=rfc4733
mailboxes=280@device
use_avpf=no
ice_support=no
media_use_received_transport=no
trust_id_inbound=yes
send_pai=yes
rtp_symmetric=yes
rewrite_contact=yes
pjsip.aor.conf
[280]
type=aor
mailboxes=280@device
max_contacts=1
remove_existing=yes
qualify_frequency=60