How to read SIP hangup causes and translate them into plain English

Every finished call leaves behind a SIP hangup cause — a code and a short reason describing how it ended. On their own the numbers look cryptic, but each one maps to a plain-English outcome that tells you whether to blame the list, the carrier, or nothing at all. Learn to read them and you stop guessing.

Where the causes are grouped

The Dial Log Report separates the calls a server placed by their SIP response code, and for each one it records the SIP hangup cause code, the unique ID, and the SIP hangup reason text. That grouping is the whole trick: instead of one call at a time you see the distribution. A list full of Hangup cause 17 reads very differently from one full of cause 21, even though both are failures.

flowchart LR
  A[SIP hangup cause] --> B[16 normal clearing]
  A --> C[17 user busy]
  A --> D[18 no answer]
  A --> E[21 call rejected]
  A --> F[34 congestion]
  A --> G[38 network out of order]
  B --> H[call ended cleanly]
  C --> H
  D --> H
  E --> I[carrier or number issue]
  F --> I
  G --> I

The common codes in plain English

• 16 Normal Clearing — the call ended the way calls normally do. By itself it is not an error; a healthy campaign is full of these.
• 17 User Busy — the line was engaged. Normal in small amounts. A spike usually means a bad list segment or a number being hammered too often by Lead recycling.
• 18 No User Responding — nobody answered in time, a no-answer. Expected on cold lists; only worrying if it crowds out everything else.
• 21 Call Rejected — the far end actively refused the call. In bulk this points at the Carrier blocking your traffic or a caller-ID that is being filtered, not a list problem.
• 34 No Circuit / Congestion — the path was full. A CONGESTION flood means a trunk or upstream capacity issue; slow your pacing and check with the carrier.
• 38 Network Out of Order — something upstream is broken. Rare and almost always on the carrier side; collect the unique IDs and report them.

Tying a cause back to a real call

The cause code tells you what kind of problem you have; the unique ID tells you exactly which call. Every row in the Dial Log carries that unique ID alongside the cause and the SIP hangup reason text. When one cause looks suspicious, grab a few of those IDs and use them to trace the same calls in the Carrier Log Report, where the raw response code from the provider sits next to it. That side-by-side view stops you from confusing a normal hang-up with a carrier reject that happens to land in the same bucket.

Pay attention to the reason text, not just the number. Two calls can both show cause 21, but one reason might read like a plain refusal and another like a block tied to your caller-ID. The text is where the carrier sometimes spells out what it objected to, and it is the difference between rotating a CID and replacing a whole trunk. The number gets you to the right neighbourhood; the reason gets you to the door.

Translate before you act: a cause-17 spike is a list to clean, a cause-21 wall is a carrier to call. Pull the unique ID from the Dial Log for any cause you want to trace end to end. For the full method see the troubleshooting playbook, and when a cause clearly points at the provider, read how to read the Carrier Log Report. If you would rather have the reporting set up and the box tuned for you, see how VICIfast runs it.