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

# Orchestration

> How Prequel Import schedules extracts and loads, and how to trigger them through the API.

![Import orchestration summary](https://storage.googleapis.com/prequel_docs/images/import-orchestration-summary.png)

Prequel Import orchestrates your pipeline automatically. Each [Dataset](https://docs.prequel.co/import/core-concepts/datasets.md) runs extracts on its own schedule, and each successful extract kicks off a load for the [Streams](https://docs.prequel.co/import/core-concepts/streams.md) that read from that Dataset.

## Scheduled extracts

Set how often a Dataset is extracted with `frequency_minutes` on the [Create dataset](https://docs.prequel.co/import/api-reference/datasets/create-dataset.md) or [Update dataset](https://docs.prequel.co/import/api-reference/datasets/update-dataset.md) request. Allowed values are `15`, `60`, `360`, `720`, and `1440` minutes; frequency availability depends on your plan. Schedules run at fixed UTC times shared across your account and daily Datasets are extracted at midnight UTC by default. Contact your Prequel representative to enable faster frequencies or change the time of day.

## Loads after an extract

When an extract succeeds, Prequel Import creates one load for each enabled Stream that reads from the Dataset and has records to deliver. If a Stream's circuit breaker is tripped, its load is recorded as skipped instead of delivered. A failed extract creates no loads.

Each load delivers the Stream's records that have not been delivered yet, plus any record whose content changed since its last delivery. For retries, redelivery, and circuit breaking on failed deliveries, see [Delivery failure handling](https://docs.prequel.co/import/features/delivery-failure-handling.md).

## Controlling orchestration

You can control orchestration of extracts and loads using the `is_enabled` flag on each resource:

* **Dataset**: Disabling a Dataset stops its scheduled extracts and the loads they start. Use [Update dataset](https://docs.prequel.co/import/api-reference/datasets/update-dataset.md).
* **Stream**: Disabling a Stream stops automatic loads to it while extracts continue for the Dataset. Use [Update stream](https://docs.prequel.co/import/api-reference/streams/update-stream.md).

## Triggering through the API

In addition to or instead of Prequel's orchestrator, you can run extracts and loads using Prequel's API. See [Initialize extract](https://docs.prequel.co/import/api-reference/extracts/initialize-extract.md) and [Initialize load](https://docs.prequel.co/import/api-reference/loads/initialize-load.md).

An extract with a `filter` does not start loads, because Prequel does not assume what a resync should deliver. Initialize a load afterward to deliver its changes.


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