Outgoing phone calls disconnect after 20 seconds

Outgoing phone call problem. All my incoming phone call work fine however when I call out the call will be disconnected after 20 seconds. The asterisk server is behind a DMZ so its nothing to do with any port forwarding. I have had the same problem when it was set up with ports forwarding. I have no problems with internal phone calls disconnecting

I have eddited out my passwords and ip address

sip.conf

Code:

[general] canreinvite=no context=default port=5060 bindaddr=0.0.0.0 nat=no externip=MYEXTIP localnet=192.168.16.0/255.255.255.0 register => absorbed:MYPASSWORD@sip.voipuser.org/448449863190 [voipuser] type=friend context=incoming_voipuser username=absorbed secret=MYPASSWORD host=voipuser.org fromuser=absorbed fromdomain=voipuser.org insecure=very qualify=no [103] type=friend context=default host=dynamic username=103 secret=MYPASSWORD callerid="Dave" <103> canreinvite=no insecure=very qualify=no

extensions.conf

Code:

[general] nat=no externip=MYEXTIP localnet=192.168.16.0/255.255.255.0 canreinvite=no [incoming_voipuser] exten => 448449863190,1,NoOp(--- ${CALLERID} calling on VoIPUser (${EXTEN}) ---) exten => 448449863190,2,Dial(SIP/103,20) exten => 448449863190,3,Answer exten => 448449863190,4,Wait,1 exten => 448449863190,5,Voicemail(u1) exten => 448449863190,6,HangUp [outgoing_voipuser] exten => _81.,1,Dial(SIP/${EXTEN:2}@voipuser,60) exten => _81.,2,Congestion [default] include => outgoing_voipuser exten => 103,1,Dial(SIP/103)

Hi.

I think the nat=no in the general section may be a problem, I assume the * interface is an internal address? This is because you have defined a local net or have you two interfaces ?

The setting is correct for the internal phones as there is no nat but unless the * its self has an external IP address then it is nat'ed

I would remove the nat=no from general and add the relevent setting to each section as required.

Ian

I have tried the multiple combinations of nat= yes or no within the individual setting however the outgoing call are still being disconnected after 20 seconds. When I do a sip debug I have seen this line

Code:

Scheduling destruction of call 'B16E7C26-4C96-4C92-A393-E49FD236AFE5@192.168.16.103' in 15000 ms

Could someone explain if this is normal and what it does?

When call is dropped after about 19 seconds the call recipient gets a engaged tone and the caller side just goes quiet but does not disconnect the call.

Hi

Ok can you confirm how you system is set up.

if the * does not have an external address and the phones are on the same subnet then set the sip.conf

in general section of sip.conf REMOVE the nat setting.

in the phone section set nat=no
in the sip trunk section set nat=yes

REMOVE the setting you have todo wit IP addresses and nat in the extensions.conf they shouldnt be there.

also add canreinvite=no to the sections in the sip.conf

finally just connect to asterisk with asterisk -vvvvvr and see what happens when a call drops. you may well get a plain english reason for the call dropping

Good luck

Ian

Ok no luck so far

My setup is asterisk server (FreeBSD 5.3) with sip phones connecting internally there is a broadband gateway to the outside world which is behind nat. So I have the [voipuser] set to nat yes and the other sip users set to nat no.

I have added canreinvite=no to each section and this still has not fixed my problem. However when I changed canreinvite=yes to both sections I could make phone calls longer than 20seconds but the sip caller can not be heard by the recipient. The asterisk console is not showing anything when the call is disconnected.

Also if my nat setting are wrong would incoming call still work because I have no problems with incoming calls.

This is what I get from the asterisk console it does not acknowledge the recipient being disconnected only when the caller hangs up. (i have edited out my phone number with x)

Code:

-- Executing Dial("SIP/103-521c", "SIP/01274xxxxxx@voipuser|60") in new stack     -- Called 01274xxxxxx@voipuser     -- SIP/voipuser-ec5b is making progress passing it to SIP/103-521c     -- SIP/voipuser-ec5b answered SIP/103-521c     -- Attempting native bridge of SIP/103-521c and SIP/voipuser-ec5b   == Spawn extension (default, 8101274xxxxxx, 1) exited non-zero on 'SIP/103-521c'     -- Got SIP response 481 "Call Leg/Transaction Does Not Exist" back from 216.127.66.119

Ok here is what I have for voipuser.
and it works fine

Code:

[voipuser] type=friend username=xxxxxx secret=xxxxxxx host=216.127.66.119 fromuser=ianplain fromdomain=216.127.66.119 nat=yes canreinvite=no insecure=very ;qualify=60000 disallow=all allow=g729 allow=alaw allow=ulaw allow=gsm

And here is what working phone

Code:

[phone17]  disallow=all  allow=alaw  allow=g729  allow=ulaw  type=friend  host=dynamic  dtmfmode=rfc2833  secret=xxxx  mailbox=2206  context=international  callerid="MITEL5215" <xxxx>  nat=no  canreinvite=no  transfer=yes  callgroup=1  pickupgroup=1

Here is the full explanation or 481
o 481-Call Leg Transaction Does Not Exist: Indicates that the server was ignoring the request of bye or cancel since there is no matching Invite transaction.

So that looks like a reinvite problem.

My other observation is that you havent defined any codecs.

Also is the external address defined definately the correct one?

basicly though Nat needs to be no for the phones yes for the trunk and no entry in the general section and canreinvite no for all including the general section.

when the call is Up what do you get if you do a "sip show channels" ?

Should look like

Quote:

216.127.66.119 0122585238 57a2350e3fb 00103/00000 alaw Tx: ACK 192.168.0.8 phone17 60a50000-18 00101/00003 alaw Rx: ACK

Try these and and see how it goes

Ian

I have had no luck so far and no one seems to of had this problem. So I rebuilt my server with fedora core 4. I was using freebsd 4.3 my config files work first time no problems at all.

thx anyway

absorbed :

I have had no luck so far and no one seems to of had this problem. So I rebuilt my server with fedora core 4. I was using freebsd 4.3 my config files work first time no problems at all. thx anyway

If anybody else has seen this in astrerisk@home 2.8 and have managed to resolve it without using the fedora core I would appreciate it if you could share it. It seems that there are now two people who have had this problem, except mine drops after about 10 seconds. canreinvite=no is set on both the extension and the trunk, I have tried to force the extension to Alaw and the trunk to ulaw, but asterisk still tries to perform a native bridge. Everything I have read says that asterisk shouldn't attempt it if canreinvite=no or the codecs are different. obviously not the case here. Thanks David

arnetnz :

If anybody else has seen this in astrerisk@home 2.8 and have managed to resolve it without using the fedora core I would appreciate it if you could share it. It seems that there are now two people who have had this problem, except mine drops after about 10 seconds. canreinvite=no is set on both the extension and the trunk, I have tried to force the extension to Alaw and the trunk to ulaw, but asterisk still tries to perform a native bridge. Everything I have read says that asterisk shouldn't attempt it if canreinvite=no or the codecs are different. obviously not the case here. Thanks David

This may or may not be related, but I had a similar problem after setting up a "diverter" rule in extensions.conf, which would accept SIP calls from one free-inbound provider (i2telecom/voicestick), check caller-ID against an OK-list, then play a dialtone, accept 10 digits and dial out through an IAX long-distance service, bridging the two streams together.

When I first set this up, everything worked great so long as dialing (at the tone) and remote pickup all happened within exactly 20 seconds... otherwise the call would drop. But, if dialing, ringback and remote party pickup happened fast enough, it would stay connected indefinitely.

The solution was just to put an "Answer" rule at the very top, before playing the dialtone or reading in any digits. Without this, apparently the call wouldn't supervise, causing i2telecom (or perhaps the originating telco?) to disconnect after 20 seconds, despite working two-way audio.

The subsequent Dial() command, to whatever number the caller keyed in would pass its completion status back to the incoming SIP provider, so that provided supervision also, but only if the call was picked up in time.

Anyway, your issue could be something else, but the "20 seconds" caught my attention. Hope this helps...

About to rez and old thread but has anyone actually found a solution to the 20 second dropout issue? It seems various people are experiencing it (still).