| Previous | Next |
| TAPI_E_INVALHOOKSWITCHDEV | TAPI_E_NOREQUESTRECIPIENT |
TAPI_E_DROPPED
The call was disconnected
TAPI_E_DROPPED is HRESULT 2147745841 (0X80040031) from Tapi3err.h. AllStat describes it as “The call was disconnected.” The result normally appears while performing call control after the provider or remote party has already disconnected the call. Its useful interpretation is precise: the disconnected call is not valid for the exact provider, object, method, or lifecycle state that consumed it.
API and object context
- This result belongs to TAPI 3 address, call-control, phone, terminal, media-provider, request, or call-center processing.
- Associate this result with one concrete native method among ITTAPI, ITAddress, ITBasicCallControl, ITCallInfo, ITStream, ITTerminal, ITPhone, provider events, and Assisted Telephony.
- Before releasing objects after this result, retain provider identity, object generation, current state, requested transition, and the event sequence surrounding performing call control after the provider or remote party has already disconnected the call.
When the application constructed the disconnected call, it usually calls for caller-side validation. When enumeration or an event supplied it, it more often points to stale lifetime, provider replacement, device removal, configuration drift, or cross-session reuse. The ownership distinction for this HRESULT prevents a broad repair from masking the real defect.
Why this condition occurs
- It can also indicate that the disconnected call belonged to another address, call, device, provider generation, user session, or COM object.
- An asynchronous event can produce it when removal, disconnect, cancellation, reinitialization, or a state transition occurs after validation.
- A wrapper can trigger it by converting an enumeration, identifier, duration, structure size, encoding, or ownership rule incorrectly.
- It can occur when the caller supplies a malformed, unsupported, stale, or out-of-scope disconnected call.
Choose among those branches only after comparing the failed disconnected call with live capabilities or objects obtained from the same provider generation. A well-formed value can still be invalid because its scope, owner, state, or lifetime no longer matches the operation.
Telemetry for a useful incident
- Record it, 0X80040031, the native method, operation ID, elapsed time, process and thread IDs, application build, architecture, and COM apartment.
- capture call state transition, disconnect code, remote address, last successful method, provider event order, and media shutdown.
- Keep provider version, address/call/phone/terminal identities, current state, requested state, object generation, and the first lower-level error that preceded it.
- Redact telephone numbers, credentials, recordings, user-user data, and directory attributes from it telemetry while retaining lengths, types, hashes, ranges, capability flags, and opaque correlation identifiers.
A useful incident can answer whether the disconnected call came from configuration, enumeration, user input, cached state, or a provider callback. The result record should also show whether the operation succeeds with a newly enumerated object or freshly validated value.
How to isolate the fault
- Capture it at the first native return and identify the exact method used for performing call control after the provider or remote party has already disconnected the call.
- Enumerate current provider capabilities or objects and compare them with the disconnected call rejected by it.
- Validate the disconnected call without silently correcting it; check type, size, range, encoding, identifier scope, ownership, and lifetime as applicable.
- Reconstruct the event timeline before it to find removal, disconnect, cancellation, service restart, provider reinitialization, or another state transition.
- Reproduce it with one object and one operation, then vary only the suspected field or state while tracking partial output and side effects.
Correction strategy
The evidence-based correction for this HRESULT is to treat the call as terminal, stop media, release call-specific objects, and create a new call for another attempt. After it follows a provider, device, service, or configuration change, obtain a fresh object because an old interface or opaque identifier may remain invalid after recovery.
Blind repetition is unsafe until asynchronous completion and possible side effects have been reconciled. User-visible handling of it should describe the actionable condition—invalid data, wrong state, missing resource, unavailable service, or policy refusal—rather than displaying only the HRESULT.
Representative scenario
For example, the remote party hangs up while an application is preparing a transfer; the transfer fails and cleanup follows the disconnect event The application should log the failed disconnected call, provider and object generation, native method, and state-changing event, then verify recovery with a newly obtained object instead of reusing the instance that returned it.
Do not confuse it with
TAPI_E_DESTNOANSWER describes setup with no answer; it means an existing call was disconnected Keep that distinction when building dashboards and exception mappings for this HRESULT; otherwise input defects, lifecycle races, capacity limits, service outages, and final call outcomes collapse into one misleading category.
Developer and administrator actions
Operational repair for this HRESULT should target only the component identified by evidence: the Telephony service, service provider, COM server, device, configuration record, application package, or remote system. Copying arbitrary DLLs or broadly deleting registration data can create additional failures without correcting the original disconnected call.
References
- Microsoft: TAPI_E constants — official Microsoft documentation used to interpret it.
- Microsoft: ITBasicCallControl
- Microsoft: TAPI 3.1 Reference
- Microsoft: Telephony API interfaces
Looking for a different code? Search another status or error code.