Every ViciDial admin has had the day where the floor is live, the calls are not working right, and the room is looking at you. Dropped calls, one-way audio, choppy voice, agents getting kicked. The hard part is not fixing it, it is knowing where the problem actually lives, because you can spend hours tuning your box when the issue is upstream. Here is how to tell the difference and when the carrier is the answer.
Separate your box from the path
Before you change anything, figure out the shape of the problem, because that tells you where to look.
If a symptom hits one agent, one headset, or one subnet while everyone else is fine, it is local. Chasing your carrier over a single agent's choppy audio wastes a day. But if the same symptom lands on every agent at once, and especially if it gets worse exactly when you raise your dial ratio, that is the fingerprint of a path or carrier problem. Issues that scale with your call volume are not coming from a single seat.
Write the pattern down before you touch a setting. Who is affected, when it started, and whether it tracks your pacing. That one observation decides whether you are tuning Asterisk or calling your carrier.
The usual suspects, in order
Most ViciDial connectivity complaints fall into a short list, and they have a natural order to check.
- One-way or no audio. The call connects, then audio fails one direction. This is almost always media path, NAT, or firewall, not the carrier's signaling. Rule out your own NAT traversal and RTP port handling first. If it survives that and follows specific routes, the carrier's media handling is in frame.
- Choppy or robotic audio. Usually jitter or packet loss somewhere in the path. Check your own bandwidth and server load, then whether it correlates with volume. Audio that degrades as the floor ramps points upstream.
- Calls dropping mid conversation. If drops scale with concurrent call count, you are hitting a capacity ceiling, either server resources or trunk capacity. We break the capacity side down in ViciDial and Asterisk trunk capacity under load.
- Long delay before ringing, or false answers. Post dial delay and false answer supervision are their own category, covered in post dial delay on outbound call centers and dead air and false answer supervision.
When the symptoms are carrier-shaped
Some problems are the carrier's by nature, no matter how well your box is tuned. If your calls connect at a poor rate, degrade when you push volume, or behave differently depending on the destination, that is routing, and routing is something only your carrier controls. A route that completes a fraction of your calls, or buckles when the floor ramps, is not a setting you can fix in ViciDial.
That is the point where the honest answer is a carrier conversation. The traits that separate a carrier built for dialer traffic from one that quietly costs you connects are laid out in our guide to choosing a termination carrier for a dialer and our carrier guide for ViciDial and Asterisk call centers. A carrier that connects your traffic by IP, holds capacity under load, and puts a real person on the line during a live shift is often the fix you were trying to find inside your server.
Good troubleshooting is mostly knowing where not to look. Separate what your box is doing from what the path is doing, check the local causes first, and when the symptoms scale with your volume, stop tuning and talk to a carrier that was built for traffic like yours.