Skip to main content

Scheduling a notebook

A scheduled notebook runs unattended at a recurring time, with date variables in your SQL and Python that resolve to fresh values on every run. A notebook that asks for "yesterday's transactions" today will ask for the previous day's transactions tomorrow, automatically.

Goal​

You finish this page with a notebook scheduled to run on a recurring schedule, with at least one date variable used inside it.

Prerequisites​

  • An existing notebook that runs successfully end-to-end when you trigger it manually. Schedules execute the notebook exactly as you saved it; if it fails interactively, it'll fail on the schedule.
  • Your account's daily execution quota covers the cadence you want (system default 100/day; admins can override).

Steps​

1. Open the notebook and trigger scheduling​

Open the notebook from the Notebooks IDE. In the menu bar at the top of the editor, click File → Schedule Execution (the label changes to Manage Schedule if a schedule already exists for the notebook).

Schedule Notebook Execution dialog

2. Fill in the schedule details​

  • Schedule Name (required) — appears in the Scheduled Tasks management view.
  • Description (optional) — shown alongside the name.
  • Schedule Type — Recurring for hourly / daily / weekly / monthly cadences, or One-time for a single future run (QRY sets a dummy cron and max_executions=1).
  • Timezone — defaults to your tenant's default. Pick the one you want delivery times to honour.
  • Frequency — Hourly, Daily, Weekly, Monthly. The form switches between hour/minute, day-of-week, day-of-month inputs accordingly.
  • Hour / Minute (or weekday / day-of-month) — the rest of the cron expression.

The dialog shows a human-readable preview at the bottom (e.g. Schedule: Daily at 09:00) so you can sanity-check before saving.

3. Use date variables in your notebook​

Date variables resolve at execution time. Use them in SQL and Python cells so the notebook always queries "fresh" data:

VariableResolves to
{{today}}The execution date (timezone-aware)
{{yesterday}}The day before the execution date
{{this_week_start}} / {{this_week_end}}Monday / Sunday of the execution week
{{this_month_start}} / {{this_month_end}}First / last day of the execution month
{{last_month_start}} / {{last_month_end}}First / last day of the previous month

Example SQL cell:

SELECT *
FROM orders
WHERE order_date >= '{{last_month_start}}'
  AND order_date <= '{{last_month_end}}'

4. Save the schedule​

Click Create Schedule. The notebook now appears in the Scheduled Tasks management view with type Notebook, the cadence you chose, and an Active status.

A green Scheduled badge appears in the notebook's menu bar. Click it any time to open the same dialog and adjust the schedule, pause it, or delete it.

Result​

The notebook runs at its next scheduled tick, with date variables resolving to the execution time.

Common issues​

The schedule preview shows the wrong time. You probably picked the wrong timezone. Re-open the dialog and switch — the cron expression rebuilds for the new timezone.

Date variables produce empty queries. The variable substitution happens at execution; in the editor, the SQL still has the literal {{today}} text and won't run as-is unless you click the cell-level "preview substitution" toggle. Check the Scheduled Tasks page → click into the notebook task → view the most recent execution log to see the actual substituted SQL.

Schedule disabled after a few runs. Three consecutive failures auto-disable a schedule. Open the task in Scheduled Tasks, look at the latest execution error, fix it, and re-enable.

One-time schedules fired too early / too late. Pick the timezone explicitly when you create the schedule. The default is the tenant default, not your browser's local time.

See also​

QRYA product of IXEN.