Keep data fresh
Esta página aún no está disponible en español. Se muestra la versión en inglés.
Put your data on a timetable: syncs first, then the projects and dashboards that read from them, so nobody reads last week’s numbers by accident.
Before you start
Section titled “Before you start”- Permissions: only a connector’s owner can schedule its syncs. Only a project’s owner sees Schedule, and a dashboard’s owners and editors see Automate.
- Data you need: a connector that has already loaded data. Uploaded files don’t sync. You replace them instead.
- Credits this uses: a scheduled sync uses 1 credit each time it syncs something. A scheduled project run, and every dashboard refresh, uses 1 credit per 10 steps that ran, at least 1. Syncs and project runs you start yourself cost nothing.
1. Know what refreshes by itself
Section titled “1. Know what refreshes by itself”Querri answers from a copy of your data, so every answer is only as fresh as the last sync. After a sync, some things catch up on their own and some wait to be run:
| What | When it updates |
|---|---|
| A source | When its connector syncs |
| A view | By itself after the sources under it sync, along with views built on that view |
| A KPI | By itself when its Measured from table rebuilds |
| A project | Only when it runs: by hand, on its schedule, or when a dashboard refreshes it |
| A dashboard | When the projects that feed it run |
The Curio Library in these docs shows the first half of the chain. Its Average Deal Size KPI is measured from the view Pipeline by stage, so in a live Library, a sync of the tables under that view rebuilds it, and then the KPI is calculated again. Curio’s own data stops at September 2025, so its numbers won’t move.
The rest of this guide sets up the part that waits: syncs, then projects, then dashboards.
2. Decide how fresh each source needs to be
Section titled “2. Decide how fresh each source needs to be”Match each connector to how often people actually look at what comes from it:
| If people look | Schedule |
|---|---|
| Throughout the day | Incremental, Every Hour, where the connector supports it |
| Each morning | Full Refresh, Every Day, early enough to finish before work starts |
| In a weekly meeting | Every Day or Every Week, before the meeting |
Why not sync everything hourly? A scheduled sync uses a credit each time it syncs something, so one hourly schedule can use up to 24 credits a day at the default rate. That’s a lot for data nobody reads before noon.
3. Schedule the connector’s syncs
Section titled “3. Schedule the connector’s syncs”Only the connector’s owner can do this.
- In the Library, right-click the connector under Connectors and choose Configure…. Admins can also click the configure icon on its card in Settings under Connectors.
- Open Schedule, then choose Full Refresh or Incremental.
- Choose how often under Repeat, and set the Start Date and Time if the form shows them. Times use your browser’s time zone.
- Click Create Schedule.
Incremental fetches only what changed, but it works only on database tables with a cursor column, and on Salesforce and NetSuite. It also can’t see rows deleted at the source. Where a connector supports it, pair an hourly incremental schedule with a daily or weekly full refresh.
Just loaded a database? The progress card after Save and Load Data offers Sync daily, a daily full refresh at 06:00.
4. Schedule projects after the syncs
Section titled “4. Schedule projects after the syncs”A project that feeds a report should run after the syncs it reads from, with time for the views to rebuild. If the syncs start at 06:00, a project at 07:00 or 08:00 gives them room.
- Open the project and click Schedule in the Data Flow toolbar.
- Pick a Start Date, a Time and Repeat: Hourly, Daily, Weekly or Monthly. Weekly runs on Monday and Monthly on the 1st, counted in UTC, so a late-evening time can land on a different day where you are.
- Click Schedule Automation.
A scheduled run re-runs the steps below a source with new data, and a run where no steps ran costs nothing.
Watch for Deliver steps. They send email when a scheduled run reaches them, and on Test Now too. Replace a weekly manual report covers that side.
5. Refresh dashboards
Section titled “5. Refresh dashboards”A dashboard’s Automate runs every project that feeds it again. Open the dashboard, click Automate, set Repeat (Daily, Weekly or Monthly) and a Time, then click Schedule Automation. Put it after the syncs too.
Pick one schedule per project, not two.
If the dashboard is what people read, automate the dashboard and leave its projects unscheduled. Otherwise a project with its own schedule, on a dashboard that’s also automated, sends its emails twice.
6. Replace uploaded files when they change
Section titled “6. Replace uploaded files when they change”An uploaded file stays exactly as it was. When there’s a new version, click ⋯ on its row in the Library and choose Replace file…, or use Versions & replace in its panel.
Check your work
Section titled “Check your work”- The connector’s schedule reads Schedule Active, with the Next Run you expect.
- A source’s panel says Healthy, and when it last synced.
- A view’s panel says Healthy, with when it was built.
- A KPI’s Measured from shows its table’s status and when it was updated.
- The project’s or dashboard’s Recent Runs shows a run after the sync time.
Troubleshooting
Section titled “Troubleshooting”A schedule stopped running. After three failed runs in a row, Querri turns the schedule off and says it was auto-disabled. Fix the cause, such as an expired sign-in or a changed password, then switch it back on.
The connector says Reauthentication needed. Its sign-in expired. Open the connector and click Reauthenticate. Don’t disconnect it, because disconnecting deletes the data it synced.
The schedule controls are locked. Only the connector’s owner can manage its schedules. Ask them to set one up.
Deleted rows still appear. Incremental syncs can’t see deletions. Add a full refresh, and check it isn’t skipping that table with Skip sources with active incremental sync.
A view didn’t update after a sync. Frozen snapshot views don’t rebuild, and a view is skipped when its inputs failed in that sync. Check the connector’s run history for failures.
The dashboard still shows old numbers. It updates only when its projects run. Check its Recent Runs, and whether the refresh ran before the sync finished.
Next steps
Section titled “Next steps”- Sync Scheduling: full and incremental schedules, cursor fields and run history.
- Sharing and automating dashboards: everything in a dashboard’s Automate dialog.
- Replace a weekly manual report: schedule a project and send its results.
- Credits: what scheduled work costs.