> ## Documentation Index
> Fetch the complete documentation index at: https://docs.hubtalk.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# FAQ: no calls or calls are stuck

> The campaign is created but nobody is called; all contacts are exhausted; a contact is stuck in progress; the campaign is cancelled.

# FAQ: no calls or calls are stuck in the queue

> [accordion] Troubleshooting order: the campaign is created but nobody is called
>
> 1. **Campaign status.** **Running** or **Scheduled**? **Scheduled** — waiting for the start hour. **Paused** — set by an operator, or an auto-pause triggered by the share of cancellations by pre-call functions. **Cancelled** right after creation — the request was rolled back; engineers will see the cause in the logs.
> 2. **The date in the campaign name.** Tomorrow's date means the request fell under the «5 minutes before the end of the day» rule. The contacts will be **Waiting** with the reason **Outside working hours** until tomorrow's start.
> 3. **Working hours and time zone.** The campaign hours are the request's `start_hour`/`ending_hour` in the organization's time zone, cut by the adapter account window and the organization's calling window. The MX client's organization must have the `America/Mexico_City` time zone, otherwise the window is shifted.
> 4. **The hold reason on contacts.** In the contact row: **Outside working hours**, **Frequency cap reached**, **Retry no earlier than …**. In **Scheduler state** — a summary by reason.
> 5. **Trunk and binding.** If the campaign is cancelled right after creation and the script summary shows **binding: auto, ambiguous** or **auto, no subgroups**, set an [explicit binding](https://docs.hubtalk.ai/v4/compat-adapter/registry.md#dialing-binding) in the script. This is the most common cause on the first launch.
> 6. **Capacity.** The campaign speed is a step derived from `speed` (8 → 5 per minute). When the trunk is overloaded, contacts return to the queue for half a minute. Pre-call functions slow down dialing.

> [accordion] All contacts are Exhausted with zero attempts
>
> The end of the campaign day came before the window opened: the window is too short, narrowed by the organization window, or the campaign end fell on an hour that has already passed. At the end of the day all **Waiting** contacts become **Exhausted**. Check the adapter account window, the organization's calling window, and the hours the client sends.

> [accordion] A contact stays Dialing for more than 10 minutes
>
> The platform returns such contacts to the queue on its own if they have no unfinished attempt. If the client withdrew the contact during the conversation, after leaving **Dialing** it becomes **Suppressed** with no retries. If contacts are stuck en masse, contact the platform engineers.

> [accordion] The campaign is cancelled, but the client keeps sending the same requests and gets «Internal error»
>
> This is by design: a completed or cancelled campaign with the same attributes does not come back to life until the end of the day. The client must change `cname` or the start hour, or wait for the next day. In V2 such a campaign would have come back to life; in V4 the operator's decision takes precedence.

> [accordion] A contact is not dialed because of Do not call, frequency or voicemail
>
> V4 rules apply to adapter campaigns the same way as to all others: the organization's **Do not call** registry, the organization's [frequency policy](https://docs.hubtalk.ai/v4/platform/campaigns/frequency.md), and the voicemail policy. The reason is in the contact row and in the scheduler state. V2 had no such rules, so this needs to be explained to the client.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.