Loading

"[SQLSTATE:58S01] There was an error during loading database: The database does not exist.: DatabaseId: "hyper.db:shadow-extract-" when refreshing an extract on Tableau Cloud involving a federation

Date de publication: Sep 1, 2026
Description

A federated (multi-connection) data source — for example a relational connection such as Databricks paired with a file connector such as Google Drive — refreshes as a single extract job in Tableau Cloud. During the refresh, each connection is pulled into its own local Hyper database and then joined via a federation step.

This error occurs when the smaller data source idles for more than 60 minutes waiting for the larger data set to extract before they can be federated. 

Résolution

This issue is currently under investigation as a defect. In the meantime, review the error with the following steps:

  1. Confirm the data source is federated (Edit Data Source → two+ connections feeding one extract).
  2. Identify the slow connection (usually the relational/warehouse leg, e.g. Databricks).
  3. Reduce that leg's load time so it re-attaches within ~3600s: narrow its query/filters, have the source-system owner tune it (filtering, limiting rows).

Alternatively, run the extract in Tableau Bridge rather than Tableau Cloud, which will not hit the same limit for shadow extract files.

Why: the error is the fast leg's shadow extract being idle-evicted before the join re-pins it. Getting the slow leg to finish inside the idle window prevents the eviction.

 

Ressources supplémentaires

Tableau Cloud — Refresh Data on a Schedule
https://help.tableau.com/current/online/en-us/schedule_add.htm

Tableau Help — Federated / Multi-Connection Data Sources (Join Data from Multiple Connections)
https://help.tableau.com/current/pro/desktop/en-us/joining_tables.htm

Numéro d’article de la base de connaissances

005393603

 
Chargement
Salesforce Help | Article