Manage API fetches — schedule, purge & connect, preview, active toggle

Do it

1. Preview — what tables did the fetch create, and which one matters?

   From the endpoint's Action menu, click Preview. You'll see a list of tables DashboardFox created from the JSON response, named like API_2_2, API_2_3 — the prefix matches the endpoint, the suffix is the table within it.

   Click any table to load its sample data. You're looking for two things:

   • Which table holds your actual data. If your JSON had a top-level values array of records, that's the main table. Other tables are usually metadata or wrapper fields — useful for context, but not what you build reports on.
   • Whether the columns and types look right. If a date column shows up as a string, the JSON didn't preserve type info — you'll handle that in App Builder. If a column you expected isn't there, the response_path may be wrong.

   Nested JSON note: if your response had nested arrays (records-with-children, or arrays-of-arrays), DashboardFox flattens them into multiple related tables, with auto-generated key columns to preserve the relationships. To use them as one logical dataset, you'll define joins between them in App Builder. The Module 3 semantic-layer lessons cover the joining patterns.

2. Schedule — automate the fetch

   From the Action menu, click Schedule. Set the frequency that matches your data's rhythm — daily for slow-changing reference data, hourly for active operational data, more often if you've coordinated with the source's rate limit.

   Save. The schedule starts running on its next interval. Verify by checking the data in App Builder after the next scheduled time has passed.

   Important limitation, current behavior: there is no "edit schedule" view. Once you save, the schedule is set — but if you open the Schedule dialog again later, you won't see what's currently configured. The dialog opens fresh every time. To change the schedule, fill in the new values and save (the new save replaces the old). To leave it alone, cancel out without saving. Edit-schedule visibility is on the product roadmap.

   To temporarily pause a schedule without losing the config: don't open Schedule. Instead, flip the endpoint's Active toggle to OFF. The schedule is preserved; it just won't fire while inactive. Flip back to ON to resume.

3. Purge & Connect — when the JSON shape changed

   Use Purge & Connect when the structure of the data you're fetching has changed in a way that affects which tables or fields exist:

   • You realized the response_path was wrong and want to re-fetch with the corrected path.
   • The source API's JSON shape changed — fields renamed, fields added or removed, structure reorganized.
   • The data is fundamentally different from what's currently stored and you want a clean slate.

   Purge & Connect deletes the existing tables and creates fresh ones from the new fetch. This is heavy-handed. If reports or App Builder configurations were built on the old tables and the field names or table names change, those reports will break. After Purge & Connect, review every report built on this endpoint's data — fix any broken field references in App Builder first, then in the affected reports.

   If the change you're making is small — adding new fields without removing existing ones — a regular Connect usually preserves your existing tables and just brings in the new fields. Try Connect first; only fall back to Purge & Connect when you need the table structure rebuilt.

4. Active toggle — pause without delete

   The Active toggle on the endpoint controls whether scheduled fetches run. ON = schedule runs as configured; OFF = schedule paused, configuration preserved.

   Use cases:

   • Source API is down for maintenance — flip to OFF, flip back when they're back up.
   • You're about to do a Purge & Connect and don't want a schedule firing in the middle of it.
   • Cost concerns or rate-limit pressure — pause non-essential endpoints temporarily.
   • You're testing changes and want to control exactly when fetches happen.

   Manual Connect still works while Active is OFF — the toggle only affects the schedule, not your ability to fetch on demand.