Tableau Cloud Help

Tableau Bridge supports extract refreshes of published data sources and embedded data sources. Publish data sources allow you to widen the audience for your data analysis within your organization. Embedded data sources are saved to a workbook and can be shared easily without having to separately share the data source.

Both published data sources and embedded data sources have advantages. In general, a published data source refresh through bridge is best for long-running refresh tasks. An embedded data source refresh through bridge is best for short running refresh tasks and for integrated data sources.

The table below shows a few recommendations for choosing between published data sources and embedded data sources based on performance when using Bridge.

Note: Performance can vary depending on the hardware and infrastructure you’re running on.

Use a published data source	Use an embedded data source
Publishing data sources is a step toward centralizing data management. You can share the data source with multiple workbooks and create policies geared toward minimizing data source proliferation and helping people find the right data for the work they do. For more information about published data sources, see Best Practices for Published Data Sources.	You don’t want to share the data sources. Data is available only inside the workbook; it isn’t available for other users to connect to.
When connecting to an integrated data source that includes only private network connections or where there are only lightweight public connections along with private network connections.	The embedded data source includes large data sets from a public network data source and short-running tasks from private network connections.
For long-running refresh tasks, and depending on your Bridge deployment, a published data source can deliver better performance.	For short running refresh tasks. Using embedded data sources may not provide optimal performance if the refresh time for your data source is expected to be more than 10 minutes on an existing published data source or workbook with direct connections to data. To ensure that long-running refresh tasks don't take up all system resources and don't prevent refreshes of other extracts on your site, Tableau Cloud enforces a timeout limit of 120 minutes for refresh tasks. See Time limit for extract refreshes.

Troubleshooting an embedded data source extract refresh

If a refresh task reaches the timeout limit, you can try to resolve the issue using the following options.

• Reduce the size of extracts. For more information, see Optimize for Extracts.

• If you’re using a full refresh, using an incremental refresh may help in some instances. For more information, see Staying within the timeout limit.

• Convert the embedded data source to a published data source.