Freeze one failed inbound attempt
Record one authorized non-emergency call with timestamp and timezone, called number, expected primary and alternate destinations, PBX addresses, SIP Call-ID where available and the caller’s observed result. Export the current SureVoIP destination order and collect edge and PBX evidence before editing DNS, firewall or routing.
The symptom “the backup never rang” does not establish that the primary timed out. A provider may have no alternate configured, may reach a responding system, or may stop after a result that the published SureVoIP policy does not treat as a failover trigger.
Prove the alternate is on the selected number
SureVoIP documents up to five destinations per number and says additional destinations are configured through support. Confirm the affected number, the exact ordered FQDN or IP destinations, optional ports, activation state and the time of the last change. Do not infer the list from another number or from a PBX failover rule.
The portal guide says an omitted destination port defaults to 5060 and displays UK numbers in international format with 44 replacing the leading zero. Preserve the number and destination exactly as configured so a formatting or selection error is not mistaken for provider failover behavior.
Classify the first destination outcome
SureVoIP says the next destination is attempted when the preceding one times out and explicitly says Busy and Unavailable do not count. Preserve DNS resolution, provider delivery address, transport result, SIP response and elapsed time for the first destination. Do not translate the guide’s plain-language labels into one assumed SIP status code without packet or provider evidence.
If the first system returned a response, identify which component produced it and whether that result is intentionally terminal under the current service policy. Changing timers or blocking responses can alter caller treatment and should not be used to simulate a timeout.
Separate network silence from an application answer
Follow the attempt from the provider-facing address to the PBX listener, then through endpoint identification and inbound routing. A firewall, SBC or PBX can receive the request and send a response even when no user device rings. Conversely, a DNS, routing or listener failure can produce silence at one observation point without proving what SureVoIP received.
Correlate the same transaction across capture points by timestamp, Call-ID, branch and tags. SureVoIP asks self-managed PBX operators to provide a SIP trace for call-setup problems; capture only the authorized interval and protect credentials and personal call data.
Verify the second attempt and destination capacity
When the first destination genuinely times out under the provider’s policy, confirm whether an INVITE reaches the next configured address and whether that system accepts the number. Check its DNS, listener, source restrictions, route match and capacity independently. A correctly triggered second attempt can still fail after delivery.
The public guide does not publish the timeout duration, retransmission schedule, attempt count, alternate capacity or return-to-primary behavior. Obtain current order-specific confirmation rather than tuning the PBX to an invented threshold.
Retest failure and recovery with rollback
In an approved window, use a non-emergency test number and a reversible method to exercise the documented timeout path. Verify the primary attempt, exact trigger, second attempt, caller treatment, identity, audio after answer and the return to primary service. Stop if the observed response or timing differs from the agreed test.
Restore the prior destination order and PBX state if acceptance checks fail. Record expected and observed results without claiming an active-call handoff, universal response-code failover or a laboratory validation. Never use 999/112 to test this workflow.
Sources & applicability
Primary references for the technical details above. Operational examples and planning checklists are VoIP.info editorial guidance.
Examples require adaptation to your topology. No live PBX or hardware testing is claimed.