All documentation

Documentation Integrations

Keep a table in sync with an API

How a connected API's tables refresh on a schedule, how rows are matched, how fresh a table is, and what you can and cannot change in it.

A table filled from a connected API is a copy of a list kept elsewhere. Sterodata refreshes it on a schedule, so the people who read it see the source's data without anyone exporting or pasting it. This page covers the resources of an API you connected under Connect an API.

The schedule

  • Each resource refreshes every 60 minutes. The schedule starts when the resource is saved, and its first run comes one interval later.
  • To fill or refresh a table now, open the connection under Settings, Integrations, go to the Resources tab and choose Sync now beside the resource. A message says it is syncing, then how many rows were synced, or why the sync failed.
  • The first sync makes the table, in the built-in database, in the folder named after the connection.

The interval cannot be changed in the app. It can be changed through Sterodata's API, per connection or per resource: to 15 minutes or more, or to 0 to stop the scheduled runs. Anything under 15 minutes is refused.

The schedule and Sync now do not use the values set in a resource's Sync parameters, and a resource with a required parameter fails there. Keep such a resource in sync with an automation's HTTP read step, which fills them.

The Sync now item in a synced table's menu in the sidebar does not run a sync: it shows a message pointing to Automations. Use Sync now on the connection's Resources tab.

How rows are matched

Each run reads the resource's whole list from the API. A row is matched to the one already in the table by its key: the column you choose in Configure schema (primary key). A row with a known key is updated, a new key adds a row. If you choose no key, Sterodata uses a column named like an id; if there is none, the first sync is refused and says which resource needs a key.

A row removed at the source is not removed from the table. Each run also reads the whole list: reading only what changed since the last run is not something you can set in the app.

How fresh it is

A synced table says, at the top, when its data last arrived. Everyone who can read the table sees it, including the people you share it with.

It readsMeaning
As of 13:05The last sync finished at that time. A sync on an earlier day shows its date too.
Sync pendingNo sync has finished yet.
Not updating since …The latest sync failed. The rows are those of the last good sync, at that time. Signed-in readers see the source's error when they point at it.
Never syncedEvery sync so far has failed.

In the sidebar, a dot beside the table gives the same news: green when the last sync worked, amber while one runs or when it got only part of the list, red when it failed, grey before the first one.

What you can change in a synced table

The connector keeps the rows as the source has them, so you cannot change its columns' values, add or delete rows, import into the table, or change the type of a column it fills or delete one. Sterodata refuses, names the connector, and tells you to make the change at the source: the next sync brings it in.

You can add your own columns to a synced table, such as notes, a status or a formula. Those are yours to edit, and the sync leaves them alone.

Stop syncing a table

Open the table's menu in the sidebar and choose Remove from sync, then Remove. Deleting the resource on the Resources tab, or the whole connection, does the same. The table stays, with its rows as they were, and becomes an ordinary table that people can edit.

To refresh at a set time of day, combine data from several calls, or write into a table you made yourself, use an automation with an HTTP read step instead.

Last updated 29 October 2026