Tableau Extracts and Live Connections

Every Tableau data source connects in one of two ways: live or extract. The choice affects speed, freshness, and the load placed on your database. Understanding both options helps you build dashboards that respond quickly and stay current.

Live Connections

A live connection queries the original data source every time you change a view. Tableau sends a request, the database answers, and the chart updates. The data always reflects the latest records.

Strengths

• Always shows current data.
• Uses the speed of a powerful database.
• Needs no extra storage on your computer.

Weaknesses

• Slows down when the database or network runs slowly.
• Stops working when the connection drops.
• Adds query load to a busy production database.

Extracts

An extract is a compressed copy of your data stored in a Hyper file. Tableau reads from this local snapshot instead of the original source. Queries run fast, and work continues offline.

Strengths

• Delivers fast results on large datasets.
• Works without a network connection.
• Protects the source database from heavy query traffic.

Weaknesses

• Shows data only up to the last refresh.
• Uses disk space.

Side-by-Side Diagram

LIVE
[Dashboard] <----> [Database]
   Each click sends a fresh query

EXTRACT
[Database] --refresh--> [Hyper File] <----> [Dashboard]
   Clicks read the stored copy

Create an Extract

1. Open the Data Source tab.
2. Select Extract in the upper-right connection area.
3. Click Edit to open extract options.
4. Click Sheet in the lower-left corner, and Tableau prompts for a save location.

Reduce Extract Size

The Edit dialog offers settings to shrink the file.

• Filters: Keep only needed rows, such as the last three years.
• Aggregate data for visible dimensions: Roll up detail rows to the level your charts use.
• Hide unused fields: Right-click fields in the Data pane and select Hide. Tableau leaves hidden fields out of the extract.

Refresh an Extract

Full Refresh

A full refresh replaces the whole extract with fresh data. The process takes longer on large tables but guarantees complete accuracy.

Incremental Refresh

An incremental refresh adds only new rows, using a field such as Order ID or Date to find them. This method saves time on big tables with append-only data.

Manual and Scheduled Refresh

Choose Data, select the data source, and click Extract, then Refresh for a manual update. Published extracts on Tableau Server or Tableau Cloud refresh on a schedule, such as every morning at 6:00.

Choose the Right Option

Situation                          | Best Choice
----------------------------------------------------
Millions of rows, slow database    | Extract
Need second-by-second updates      | Live
Offline presentation               | Extract
Small table, fast database        | Live
Heavy use by many viewers          | Extract

Switch Between the Two

Switching takes one click. Select Live or Extract in the Data Source tab. Calculated fields, groups, and formatting stay in place. Test the dashboard after switching because some functions behave differently between sources.

Good Habits

• Pick extracts for dashboards with many viewers.
• Schedule refreshes outside business hours.
• Document the refresh time on the dashboard so readers know the data age.

Summary

Live connections deliver current data directly from the source. Extracts store a fast local snapshot that needs refreshing. Filter extracts and aggregate visible dimensions to keep them lean. Match the connection type to your speed needs, data volume, and freshness requirements.