VoIP.info Troubleshooting

Orange Business Talk site, trunk, routing or media is wrong

Find the first divergence across ordered site state, portal activation, DNS, time, certificate, eSBC guide, SIP routing and negotiated media.

Reviewed 2026-10-25Orange Business Talk product, service-document index, site-management assistance, certified eSBC and ecosystem guides and enterprise-voice material reviewed October 25, 2026; the selected country, order, site, access model, project documents and certified version govern, while the October 25, 2024 Patton guide remains guide-scoped evidence; no Orange account, order, portal, site, trunk, circuit, DNS answer, clock, certificate, SBC, PBX, number, route, packet, call, fax, media stream, failover event, emergency result, price, availability, compatibility or support response was inspected or testedAdvancedDocumentation based · not lab tested

Freeze one attempt and identify the exact service site

Record UTC and local time, country, legal service name, site, access model, number, direction, current order and project document identifiers, selected eSBC and PBX objects, expected route, answer and media result. Redact credentials, personal data and full internal addresses.

Classify the first unexpected state: portal change not active, DNS or clock wrong, certificate or TLS failure, request missing at the Orange-facing edge, number or identity normalization wrong, unexpected reroute, PBX route wrong, or signaling complete without usable media.

Compare requested, saved, activated and loaded state

Orange Business assistance separates site, voice-channel and trunk, network, call-configuration, address and advanced settings. Capture the requested change, portal value, activation or notification state and the effective eSBC and PBX configuration. A saved screen is not a loaded-state test.

Confirm that the guide matches the actual eSBC model, release, service variant and access architecture. Do not import a Patton, Ribbon, Teams, Cisco or another ecosystem template into a different owner without a current project record.

Trace DNS, time, certificate and transport in order

Preserve the DNS query and accepted answer, selected address and port, route, source tuple, clock source and offset, certificate chain, service identity, TLS alert or completed handshake and the first SIP request. A correct DNS answer can still lead to the wrong route or certificate identity, and a valid clock does not prove trust or authorization.

Use the exact project firewall and peer record. Do not open every address or port from a historical guide, disable verification broadly or treat a TCP connection as a successful trunk.

Follow number, identity, probing and rerouting evidence

For the same Call-ID and time window, preserve Request-URI, From, To, P-Asserted-Identity, Privacy, Diversion, Via, Contact, source and destination tuples, every response and the next attempted destination. Compare the number form and response treatment with the current project specification.

OPTIONS proves only the probed transaction. A second attempt proves only that an alternate path was tried. It does not establish why the first path failed, that all new calls use the intended route or that established calls survive a failure.

Separate answer, early media, DTMF, fax and two-way audio

Preserve both SDP bodies, chosen codec and packet rate, early-media headers and packets, DTMF method and events, RTP or SRTP tuples and decoded direction at authorized points. A 200 response, TLS signaling or encrypted media offer does not prove audible two-way media.

For fax, confirm that the ordered service and current project support the workflow before testing an authorized destination. Record T.38 offer, acceptance, UDP behavior and completed pages separately from voice registration and ordinary call answer.

Retest one correction and keep emergency acceptance separate

Change only the demonstrated owner, repeat the affected direction and an unaffected control, and preserve portal effective state, edge receipt, PBX route, answer, media and normal hangup. If a failure route was involved, restore the primary and prove the next call returns to it. Remove temporary logging and revert unsuccessful experiments.

Do not place an uncoordinated emergency call. Confirm country eligibility, registered site address, effective update, outbound identity, route, callback and alternate method through the provider-approved process. No Orange account, site, trunk, route, packet, call, media, failover or emergency result was inspected or tested for this diagnostic.

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.