Not connection to dialer + agent don't hear 1 way audio

[EXviciman]

Hello,

Vicibox 8.0.1 from http://vicibox.org/server/index.html | Vicidial 2.14-675a Build 180520-1749 | Asterisk 11.25.3 | Single Server (OpenSUSE 42.3) | No Digium/Sangoma Hardware | No Extra Software After Installation

We have a few weird issues with a few different providers.

1. On provider X: When we call manually everything works fine. Once we dial to the same people using the dialer the call calls through but we get a sip-silence message on asterisk cli (Client picks up & after 6 seconds phone goes down). Also call from client doesn't show on the conference panel.

2. On Provider Y: When calling using even using manual calls (either from a web client or from the agent interface) we get a one way audio - where the client can hear the agent , but the agent cannot hear a thing. I sniffed the external firewall & we actually saw that traffic goes only outside of our server ,but not to inbound to our server (Maybe besides an icmp ping every now and then). EDIT: Note I have opened all ports on external firewall + on PC firewall

The weird thing is that on Provider Z everything is working smoothly! Which leads me to suspect of a config issue....

If anyone has some insight about this it will be very helpful !

[Acidshock]

Do you have your externip set in /etc/asterisk/sip.conf?

[EXviciman]

Yes, we have

[williamconley]

1) Post your carrier configuration including (but not limited to) the account entry and dialplan entry. Mask any IP/domain or pass with "xx.xx.xx.xx"/"xx.com" or "XXXXXX".

2) One-way audio is almost always firewall-related. Test again with NO firewall for long enough to demonstrate this, then turn the firewall back on and find out which IP for the Carrier is blocked and unblock it. If you are using NAT (local/private network with shared public IP), you may wish to consider adding a public IP for your Vicidial server. It makes situations like this moderately easier.

3) If you've changed the IP of the server (local or public or both), revisit that externip setting and ensure it really does have the external IP. To be sure:

Code: Select all

curl ip.whowebwhere.com

[EXviciman]

1)
[carrier]
username=****
type=peer
secret=****
qualify=yes
host=***
directmedia=no
insecure=invite

exten => _X.,1,AGI(agi://127.0.0.1:4577/call_log)
exten => _X.,2,Dial(SIP/carrier/${EXTEN},,To)
exten => _X.,3,Hangup()

Also what do you mean by: 'Mask any IP/domain or pass with "xx.xx.xx.xx"/"xx.com" or "XXXXXX" ' ?

2) It is a little bit of a problem to shut the firewall down. We use an external ip for the server and well other providers reach it(curl ip.whowebwhere.com show correct ip) . I will further check to see about blocks though - because that makes sense.

[EXviciman]

UPDATE:

Here is the asterisk cli for the issue no.1 , I just now tried twilio as well as provider and they give me the the same error.

After 6 seconds phone from when client answers the call is hangs up + not routed to the conference.


[Jul 17 22:17:48] -- SIP/twilio-00000007 answered Local/{####DIALED PHONE NUM####}@default-00000004;2
[Jul 17 22:17:48] > Channel Local/{####DIALED PHONE NUM####}@default-00000004;1 was answered
[Jul 17 22:17:48] -- Executing [8368@default:1] Playback("Local/{####DIALED PHONE NUM####}@default-00000004;1", "sip-silence") in new stack
[Jul 17 22:17:48] -- <Local/{####DIALED PHONE NUM####}@default-00000004;1> Playing 'sip-silence.gsm' (language 'en')
[Jul 17 22:17:48] -- Executing [8368@default:2] AGI("Local/{####DIALED PHONE NUM####}@default-00000004;1", "agi://127.0.0.1:4577/call_log") in new stack
[Jul 17 22:17:48] -- AGI Script Executing Application: (EXEC) Options: (Set(_CAMPCUST=TEST5))
[Jul 17 22:17:48] -- <Local/{####DIALED PHONE NUM####}@default-00000004;1>AGI Script agi://127.0.0.1:4577/call_log completed, returning 0
[Jul 17 22:17:48] -- Executing [8368@default:3] AGI("Local/{####DIALED PHONE NUM####}@default-00000004;1", "agi-VDAD_ALL_outbound.agi,NORMAL-----LB") in new stack
[Jul 17 22:17:48] -- Launched AGI Script /usr/share/asterisk/agi-bin/agi-VDAD_ALL_outbound.agi
[Jul 17 22:17:49] == Manager 'sendcron' logged off from 127.0.0.1
[Jul 17 22:17:50] -- <Local/{####DIALED PHONE NUM####}@default-00000004;1>AGI Script agi-VDAD_ALL_outbound.agi completed, returning 0
[Jul 17 22:17:50] -- Executing [8368@default:4] AGI("Local/{####DIALED PHONE NUM####}@default-00000004;1", "agi-VDAD_ALL_outbound.agi,NORMAL-----LB") in new stack
[Jul 17 22:17:50] -- Launched AGI Script /usr/share/asterisk/agi-bin/agi-VDAD_ALL_outbound.agi
[Jul 17 22:17:51] -- <Local/{####DIALED PHONE NUM####}@default-00000004;1>AGI Script agi-VDAD_ALL_outbound.agi completed, returning 0
[Jul 17 22:17:51] -- Executing [8368@default:5] Hangup("Local/{####DIALED PHONE NUM####}@default-00000004;1", "") in new stack
[Jul 17 22:17:51] == Spawn extension (default, 8368, 5) exited non-zero on 'Local/{####DIALED PHONE NUM####}@default-00000004;1'
[Jul 17 22:17:51] -- Executing [h@default:1] AGI("Local/{####DIALED PHONE NUM####}@default-00000004;1", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----16---------------") in new stack
[Jul 17 22:17:52] -- <Local/{####DIALED PHONE NUM####}@default-00000004;1>AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
[Jul 17 22:17:52] -- Executing [h@default:1] AGI("Local/{####DIALED PHONE NUM####}@default-00000004;2", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----16-----ANSWER-----10-----4") in new stack
[Jul 17 22:17:53] -- <Local/{####DIALED PHONE NUM####}@default-00000004;2>AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... --10-----4 completed, returning 0
[Jul 17 22:17:53] Scheduling destruction of SIP dialog '55052c867882c7a82d1e5c8c368b62e9@{####MY-EXTERNAL-IP####}:5060' in 6400 ms (Method: INVITE)
[Jul 17 22:17:53] set_destination: Parsing <sip:54.172.60.1:5060;lr;ftag=as7bcd93c9;twnat=sip:{####MY-EXTERNAL-IP####}:5060> for address/port to send to
[Jul 17 22:17:53] set_destination: set destination to 54.172.60.1:5060
[Jul 17 22:17:53] Reliably Transmitting (NAT) to 54.172.60.1:5060:
[Jul 17 22:17:53] BYE sip:172.18.17.101:5060 SIP/2.0
[Jul 17 22:17:53] Via: SIP/2.0/UDP {####MY-EXTERNAL-IP####}:5060;branch=z9hG4bK0746c41d;rport
[Jul 17 22:17:53] Route: <sip:54.172.60.1:5060;lr;ftag=as7bcd93c9;twnat=sip:{####MY-EXTERNAL-IP####}:5060>
[Jul 17 22:17:53] Max-Forwards: 70
[Jul 17 22:17:53] From: "V7172217420000021811" <sip:{###CID-NUM###}@{####MY-EXTERNAL-IP####}>;tag=as7bcd93c9
[Jul 17 22:17:53] To: <sip:+{####DIALED PHONE NUM####}@mydomain.pstn.us1.twilio.com>;tag=94170020_6772d868_05f6c909-aa33-4f27-9cd2-187d29875573
[Jul 17 22:17:53] Call-ID: 55052c867882c7a82d1e5c8c368b62e9@{####MY-EXTERNAL-IP####}:5060
[Jul 17 22:17:53] CSeq: 104 BYE
[Jul 17 22:17:53] User-Agent: Asterisk PBX 11.25.3-vici
[Jul 17 22:17:53] Proxy-Authorization: Digest username="USER", realm="sip.twilio.com", algorithm=MD5, uri="sip:172.18.17.101:5060", nonce="P4G8H_ZrKOc2i9fLmsXD--mLBrdWFDiX1clARuDFp_jkkcNO", response="108425274d12b752377c1f99576b6de4", opaque="d94648793ad62a4e1b1a3a5e723a5777", qop=auth, cnonce="64a30ba0", nc=00000002
[Jul 17 22:17:53] X-Asterisk-HangupCause: Normal Clearing
[Jul 17 22:17:53] X-Asterisk-HangupCauseCode: 16
[Jul 17 22:17:53] Content-Length: 0

[williamconley]

Vicidial scripts will not transfer a call to an agent unless there is two-way sound. Firewalls block two-way sound unless all audio IPs are open and routed to the server. Note that SIP uses port 5060 ONLY for control. The audio is on a random port usually between 10k-25k udp. And the audio may arrived from an IP other than the IP of the server controlling the call. So you may get a control packet on port 5060 from one IP initiating the call, and then audio on port 15233 (randomly chosen) from a completely different IP. If your firewall is properly configured, it will route that audio packet to the Vicidial server. If your firewall is DOWN it will also route that audio packet to the vicidial server. But if your firewall blocks that packet, audio is impossible, and the call will NOT transfer to an agent and will be terminated in a pre-determined time (anywhere from 5-62 seconds).

Proper configuration of the firewall can be achieved with a "sip algorithm" that recognized the dance between the control packet and audio packet OR by having only one server getting internet packets from the public IP, thus it's a matter of "block or not block" instead of "who gets this packet?"

If you have multiple devices internally all sharing the same public IP, the SIP algorithm in the firewall is required to inspect the packets and route them properly OR you must find another method to route all these packets to your Vicidial server. For instance: Routing all traffic to port 5060 (UDP) and ports 10k-25k (UDP) to the Vicidial server is a bit of a "duct tape" method but has been known to work. There are also some "trigger" algorithms in certain routers that allow "hit" on one port to "trigger" listening on another port or port range. Can get rather complicated, and is implemented differently by different router companies.

Thus having a dedicated public IP for each Asterisk PBX is often simpler. Then you just use whitelisting (with or without Dynamic Good Guys firewall) to secure the server so it can only communicate with Carrier IPs instead of using a router.

[Flashhunter]

check the allow codec on your provider.

[williamconley]

check the allow codec on your provider.

If a codec mismatch occurs, the call doesn't die after 6 seconds: It dies immediately as a result of the failed handshake. Good reflex, though.

[Acidshock]

What firewall are you using? Check if SIP-ALG is enabled, or disabled. Flip it to the opposite. So if its enabled try disabled, and vice versa.

Also can you do a sip debug on the carrier? Let's see what ip is designated to receive media.

Another thing to check...is the software firewall enabled on the server?