Outbound calls failing at the carrier: first checks

When outbound calls fail at the Carrier, run a quick first-checks pass before you dig deep: is the Trunk active, is the peer reachable, is the dial pattern and strip correct, is the caller ID valid, and what SIP response code is the carrier actually returning? Four times out of five the answer is in those five questions.

Run the decision tree

Work top to bottom. Each check rules out a whole class of problem before you spend time on the next.

flowchart TD
  A[Outbound call fails] --> B{Trunk Active=Y?}
  B -->|No| C[Set Active=Y, rebuild conf]
  B -->|Yes| D{Peer reachable?}
  D -->|No| E[Fix register or IP auth]
  D -->|Yes| F{Dial pattern and strip correct?}
  F -->|No| G[Fix dialplan exten and strip]
  F -->|Yes| H{Caller ID valid?}
  H -->|No| I[Set a valid outbound caller ID]
  H -->|Yes| J[Read SIP response: 503 or 603]

The five first checks

• Trunk active: if the carrier's Active flag is N it gets left out of the conf files entirely, so nothing dials. Set it to Y and wait a minute for the rebuild.
• Peer reachable: run sip show peers. If the peer is unreachable, no call can leave, so fix registration or IP authorization first.
• Dial pattern and strip: confirm the dialplan exten match and the digit strip produce the number format the carrier expects, often a clean E.164-style number.
• Caller ID: many carriers reject calls with a blank or unauthorized caller ID. Make sure your outbound caller ID is one the carrier will accept.
• SIP response: read what the carrier actually returns, because the response code tells you who is rejecting and why.

On the dial pattern, the format usually needs to be E.164 or close to it. If your strip leaves the campaign dial prefix on the front, or drops a country code the carrier wants, the carrier sees a number it cannot route and rejects the call before it ever rings. The leading 9 in a typical dialplan is the campaign dial prefix and should be stripped, while the country code may need to stay depending on what the carrier expects. Get the strip wrong by one digit and every call fails the same way.

It is worth confirming the failure is consistent before you blame any single check. If every call to every number fails, the trunk, peer, or dial pattern is the likely culprit. If only some destinations fail, the carrier may simply not have a route to those numbers, which is a conversation with the carrier rather than a config change on your side.

Read the SIP response

The carrier's response code is the most useful clue. A SIP 503 Service Unavailable service-unavailable points at the carrier being overloaded or refusing your traffic, often a capacity or account issue on their side. A 603 Decline means the carrier deliberately rejected the call, which usually traces back to caller ID, an unauthorized number format, or a blocked destination. A CHANUNAVAILABLE or congestion result points back at the trunk or channel availability rather than the number.

Once you have isolated the layer, the fix is usually small. These first checks sit alongside the rest of the carrier troubleshooting in the VICIdial carrier integration guide, and the active-flag detail is covered in the carrier active toggle.

On a VICIfast server the dialer and Asterisk are already tuned and the box has a fixed public IP, so when a call fails the carrier side is usually the only variable left to check. Your dialer is live over HTTPS in under 40 seconds.