Define the Access Mode for a Dataset

You can determine how a dashboard accesses data in a specific dataset. The following data access modes are available:

Connect Live: Retrieves dataset results directly from a data source. Dragging dataset objects onto visualizations using Connect Live mode can be slower than when using In-Memory mode, since the system retrieves dataset results before they appear. However, the performance of a large dataset can improve by retrieving data incrementally.
In-Memory: Retrieves dataset results from a data source and stores them in memory. Visualizations in the dashboard that contain data from the dataset display subsets of these stored results, which can increase the speed in which data appears.

The dataset access method also affects data retrieval when you open the same dashboard in Web.

If the data source uses In-Memory access, the system creates an in-memory cube from the data when you upload the dashboard from Web. This cube is accessed when the dashboard opens in Web. If a user refreshes the data, the MicroStrategy server connects directly to the data source to retrieve new data to repopulate the cube.
If the data source uses Connect Live access, the dashboard does not contain the data. When the dashboard is accessed in Web, the server connects directly to the data source to obtain the data.

Certain data sources support only Connect Live or In-Memory mode. The dataset must support both modes to allow access mode selection. The table below contains a list of data source types organized by the access modes they support.

Direct Data Access Mode Only	Both Direct Data Access and In-Memory Modes	In-Memory Mode Only
Objects from a connected server	Relational databases, including: Picking a table Building a SQL query Typing a query Accessing Hadoop files Google BigQuery	All other data sources

Direct Data Access Mode Only

Both Direct Data Access and In-Memory Modes

In-Memory Mode Only

Objects from a connected server

Relational databases, including:

Picking a table
Building a SQL query
Typing a query
Accessing Hadoop files
Google BigQuery

All other data sources

The following rules apply to live-to-schema datasets:

Only one live-to-schema dataset can be added per dashboard
Filters cannot be added to a live-to-schema dataset
Live-to-schema datasets cannot be changed to in-memory datasets and vice versa
Objects are only executed once they are in the visualization, as such, any object errors are contained in the visualization.

The dataset must support both Live Connect and In-Memory data access modes. Toggle Live Connect Only/All Connectors on the Data Sources dialog to display All Connectors. This allows you to select from data sources that support both Live Connect and In-Memory access.