Choosing a Data Access Type for Dataset Tables

In Oracle Analytics, choosing a dataset table's Data Access setting determines whether data is cached or loaded directly from the data source.

Data Access settings:

• Live - A table gets its data directly from the data source. The source system manages the table's data source queries. This option is useful when the data is stored in a high-performance system such as Oracle Autonomous Data Warehouse. It also ensures that the latest data is used.
• Automatic Caching - A table loads its data from the cache where possible. This option can provide faster performance when you refresh a table's data from the Transform editor or from a workbook. Selecting this option causes the Reload menu option to display at the table and dataset level.

  If a dataset load fails or columns are added, data access temporarily switches to Live mode until a new reload is successful.

  This option is limited to 2GB of data after compression. If the data is larger than 2 GB or if the data takes too long to load, the data is accessed using a live query (if the connection type supports it). If you want to use automatic caching and the data is larger than 2 GB, then add column filters to the table, for example limit a date range to reduce the size of the data.

• Extracted - A table always loads its data from the cache. Extracted mode enables authors to ensure that in-memory caching of datasets is completed before allowing queries to provide consistent performance for workbooks.

  Extracted mode forces the reload job to complete before letting other users use the dataset, which ensures that it's fully cached. In addition, a dataset author can make modifications such as adding a column, and the changes are only visible to users when all of the data is available in the cache.

  If a dataset reload fails or new columns are added, workbooks use the last data loaded.

Tip: For best performance, if you have a dataset with multiple tables, set all tables to access data in the same way. Although you can can configure some tables to use caching and some to use live data, this can affect performance. For example, if the reloading of one table fails, then any tables set to use automatic caching are switched to use live data.