Refresh
How dashboard data stays current.
Refresh controls how often Flitch rebuilds the cached snapshot of a dataset. Refresh only runs at the interval you set; nothing happens in between.
Off vs on
| State | What it does |
|---|---|
| Off (default) | Every dashboard load queries the source directly. Always current, but slower as the data grows. |
| On | Scheduler refreshes at your interval and writes to cache. Dashboards serve from cache between runs. Faster page loads, fewer warehouse hits. |
Set a schedule
Per dataset
Two entry points:
- When adding a source or editing a connection on the Data page. Look for Auto-refresh data.
- From an existing dashboard, open Settings → Data Sources. Each dataset has its own toggle.
Schedules persist on the dataset, so changes apply everywhere the dataset is used.
Dashboard-wide
If a dashboard uses several datasets and you want them all on the same cadence, open the refresh dialog from:
- The prompt input when creating a dashboard ( icon).
- Or Settings → Data Sources on an existing dashboard.
Flip the Bulk auto-refresh toggle and set one schedule. It applies the same interval to every dataset the dashboard uses.
Intervals
Pick a value in minutes, hours, days, or weeks. How fast you can go depends on your plan:
| Plan | Fastest interval |
|---|---|
| Free | Every 12 hours |
| Pro and Team | Every 1 minute |
Pick something faster than your plan allows and Flitch slows it to the limit, with a note in the dialog. The limit is applied when the schedule is saved and again each time the next run is scheduled, so existing schedules slow down on their own if a team moves to Free. Slower intervals are unrestricted on every plan.
See Billing & Usage for what else changes by plan.
Manual refresh
In the refresh dialog, click Refresh next to any dataset to run it immediately, regardless of schedule.
Failures
- Reconnect required: tokens expired or revoked. Open the connection, click Reconnect.
- Auto-disabled: too many consecutive failures. Re-enable from cache settings after fixing the underlying issue.
- Error: detail shown next to the dataset. Usually a removed column or table.
Cost
Frequent refreshes mean more queries against your source, which may incur costs depending on your provider.
Match the cadence to how stale the dashboard can be. Default to daily and speed up if needed.