VICIdial “No one is in your session” Error: 8 Checks to Fix It

Short answer: “No one is in your session: 8600051” (older versions: “Noone is in your session”) means VICIdial placed the agent’s session in a conference but the agent’s phone never joined it. Check, in order: the VICIdial screen processes are running (screen -ls), the agent’s phone is registered and assigned to the right server, the conference numbers exist for that server, MeetMe has a DAHDI timing source, and for ViciPhone, a valid SSL certificate and a reachable WebSocket on port 8089.

These checks come from the official ViciBox documentation and fixes posted on the VICIdial forum (linked below), reviewed on 6 October 2026; not yet run on our lab servers, which do not run VICIdial.

What the message means

When an agent logs in to the agent screen, VICIdial gives the session a conference number (8600051, 8600052 and so on) on the telephony server and calls the agent’s phone into it. The agent should hear a prompt that they are the only person in the conference. If that call never connects, the screen reports that nobody is in the session and offers to go back or call the agent again.

So the error is almost never about the conference itself. It is about the call to the agent’s phone: it was not placed, not answered, or answered but not bridged into the conference. The checks below follow that chain.

1. Are the VICIdial processes running?

VICIdial runs its background scripts in screen sessions. According to the ViciBox Express docs, a healthy server shows 11 sockets after boot (it can take up to 5 minutes):

screen -ls

If you see “No Sockets found”, nothing is placing the calls. On the VICIdial forum, the fix offered for that case was to start Asterisk and the keepalive script by hand, then find out why they did not start at boot:

/usr/share/astguiclient/start_asterisk_boot.pl
/usr/share/astguiclient/ADMIN_keepalive_ALL.pl

ADMIN_keepalive_ALL.pl is normally scheduled in root’s crontab; check the entry exists with crontab -l | grep keepalive. If screen -ls shows many duplicate asterisk sessions, something is starting Asterisk repeatedly (for example a boot script added to cron by mistake); clean those up before testing again.

2. Is the agent phone registered?

The agent must log in with a phone that is actually online. Connect to Asterisk and look for the phone’s extension:

asterisk -r
sip show peers        # chan_sip systems
pjsip show contacts   # PJSIP systems

If the phone is unregistered or unreachable, fix that first: credentials, NAT, firewall. A phone that registers but drops after a minute or two usually has a NAT keepalive problem; our unreachable endpoint guide covers keepalives, and Disable SIP ALG covers router interference.

Then watch the CLI while the agent logs in. You should see VICIdial originate a call to the phone’s extension. No call at all points back to check 1 or check 3; a call that fails points to the phone or its network.

3. Is the phone on the right server?

In a cluster, each phone record (Admin > Phones) has a Server IP. VICIdial only calls phones assigned to the telephony server where the agent’s session lives. A phone set to the wrong server IP, or a server whose IP changed, gives exactly this error. If the server’s public IP changed, the ViciBox docs say to run vicibox-externip and update the External Server IP under Admin > Servers.

4. Do the conferences exist?

A forum reply pointed to this Asterisk warning in the log:

WARNING[...]: app_meetme.c:5258 admin_exec: Conference number '8600051' not found!

Check Admin > Conferences and VICIDIAL Conferences in the admin interface: the 8600xxx conference entries must exist with the IP of the telephony server you use. Missing entries are common after changing a server’s IP or adding a new dialer. Another experienced user noted that the same log line can also appear harmlessly when the “Agent Login Kick-all” system setting kicks participants on login, so treat it as a clue, not proof.

5. Does MeetMe have a timing source?

VICIdial’s classic conferences use Asterisk’s MeetMe application, which relies on DAHDI for timing. In one forum case on ViciBox the agent got this error until the DAHDI kernel module was loaded:

lsmod | grep dahdi
modprobe dahdi
dahdi_cfg -vvv

If dahdi_cfg reports Unable to open master device '/dev/dahdi/ctl', the module is not loaded. After loading it, the agent in that thread could join with “Call Agent Webphone” and heard the conference prompt. Make sure the module loads at boot, not just once by hand.

6. ViciPhone and WebRTC agents

With ViciPhone, the “phone” is a WebRTC client in the browser, so the failure is often on the web side. Look for “Reg. failed” in the ViciPhone panel. Then check:

• SSL certificate: ViciPhone needs a valid certificate matching the server’s FQDN. ViciBox’s vicibox-ssl sets up Let’s Encrypt for Apache and Asterisk. Self-signed certificates were a recurring cause in forum threads.
• WebSocket: per the ViciBox docs, WSS runs on TCP 8089 and must be reachable by agents. Confirm Asterisk loaded the certificate and the /ws URI: asterisk -rx "http show status".
• Certificate paths in Asterisk’s http.conf and in the WebRTC phone template must point at the real certificate and key.
• Browser microphone permission and an HTTPS agent page.

One forum fix concerned the ViciPhone settings container (VICIPHONE_SETTINGS): changing dialRegExten : 1 to dialRegExten : 0 helped when the phone was not in the default context. The poster noted that when a template is used, the template’s context wins over the phone’s own context.

For webphones that do not join automatically after login, a forum post describes enabling options.php in the agent web folder (on ViciBox /srv/www/htdocs/agc/; copy options-example.php to options.php if needed) and setting:

$webphone_call_seconds = '5';

which calls the agent webphone a few seconds after login. It needs a recent SVN version.

7-8. Network and time problems

• NAT and one-way audio. If the agent phone rings and answers but the agent hears nothing, or the session still shows empty, check RTP: UDP 10000-20000 open on the server, External IP correct, SIP ALG off at the agent’s router. See our one-way audio guide.
• Firewall whitelist. In ViciBox White List or Dynamic Portal mode, a remote agent whose IP is not in the external zone cannot register or reach WSS. Check the IP list or have the agent log in to the portal first.
• Time synchronization. Some agents see “There is a time synchronization problem with your system” alongside this error. Make sure all servers in the cluster keep time with NTP.

Check that it is fixed

1. screen -ls shows the expected sockets.
2. The agent’s phone shows as registered in the Asterisk CLI.
3. On agent login, the phone rings (or ViciPhone connects) and the agent hears the “only person in this conference” prompt.
4. The agent can go READY and receive a test call from a campaign.
5. No Conference number ... not found warnings appear in the Asterisk log during login.

Official documentation: ViciBox 12 documentation · VICIdial forum: ViciPhone “No one is in your session” · VICIdial forum: “No one is in your session” (conferences, SSL)

Related: VoIP One-Way Audio Fix: 6 Checks for NAT and RTP · Disable SIP ALG: 7 Router Fixes for One-Way Audio and Dropped Calls · PJSIP Behind NAT: Asterisk and FreePBX Settings for Two-Way Audio · Troubleshoot SIP Calls with sngrep and the Asterisk PJSIP Logger · SSL Certificate Checker

See also: Install VICIdial with ViciBox 12: Express Setup, SSL and Firewall

Frequently asked questions