Testing Tableau Dashboard Performance

What it means: the query takes a long time to run. This can be because of poorly written queries, too much data, or too many level of detail (LOD) expressions.

How to resolve:

• Check to make sure you are using initial SQL instead of custom SQL
• Re-write the query
• Bring LOD expressions to the data layer (or use Tableau Prep)

What it means: the layout of the dashboard is too complex or there are too many worksheets.

How to resolve

• Reduce the number of worksheets

What it means: this can be due to network issues or issues with the database server.

How to resolve:

• Check your internet connection
• Talk to your database administrator (DBA)

What it means: indicates the query generated is too complex. Common causes of too complex queries are too many filters or too many calculations.

How to resolve:

• Move calculations to the data layer (or use Tableau Prep)
• Change filters – reduce the number of filters or convert to action filters

What it means: there is a lot of data or the connection type could be changed.

How to resolve:

• Use a data source or context filter to reduce the amount of data
• Convert the connection type (if applicable)

What it means: data source connections bring all columns and fields from the query or data source into the workbook – often more columns than we need. More data means it takes longer to generate an extract.

How to resolve:

• Use a data source filter
• Hide unused fields

What it means: for certain data (city, state), Tableau will recognize this data as geographical, convert the data type, and assign latitude/longitude. If there are many records brought in at a granular level (for example, millions of rows of data at a street level address), this impacts performance.

How to resolve:

• Visualizations should start aggregated at a higher level (less data) and then granular (more data)