how to run previous day pipeline in airbyte using cron

Airbyte supports CRON-based schedules for each connection, but the docs only describe when to run a sync, not how to parameterize it as “previous day.” So you can schedule daily runs, but selecting “yesterday’s data” must be handled by your source configuration (e.g., an incremental cursor) or by your downstream logic; the scheduling layer itself is just time-based. [[Sync schedules](https://docs.airbyte.com/platform/using-airbyte/core-concepts/sync-schedules)] To run a pipeline once per day (e.g., at 01:00 every day) using CRON: 1. Edit your connection. 2. Under Schedule type, choose Cron. [[Sync schedules](https://docs.airbyte.com/platform/using-airbyte/core-concepts/sync-schedules)] 3. Enter a CRON expression and timezone. Airbyte uses Quartz-style CRON. [[Sync schedules](https://docs.airbyte.com/platform/using-airbyte/core-concepts/sync-schedules)] Examples (Quartz format): - Every day at 0100

0 0 1 * * ?

- Every day at 0005

0 5 0 * * ?

These will trigger the connection once per day; if your source is incremental (e.g., based on an updated_at or date column), it will naturally pick up “previous day” changes on each run. If you instead want to trigger a sync from an external scheduler (e.g., a cron job on your server) at a specific time, you can: - Set the connection to Manual schedule. [[Sync schedules](https://docs.airbyte.com/platform/using-airbyte/core-concepts/sync-schedules)] - Use your own cron to call the Airbyte API

/v1/connections/sync

with the connection ID at the desired time, similar to:

bash   0 1 * * * curl -X POST <http://localhost:8000/api/v1/connections/sync> \     -H "Content-Type: application/json" \     -d '{"connectionId": "<your-connection-id>"}'

This pattern is described in a community example for weekly syncs and works the same for daily. [[Custom weekly sync](https://discuss.airbyte.io/t/379)] Because the docs don’t show a built-in “previous day” flag, you’ll need to rely on your source’s incremental configuration or your own logic to ensure that each daily run processes the prior day’s data.