Data upgrade tasks for Cognos Analytics version 11.1

To support the optimized user experience in dashboards, explorations, and other components, and to improve query performance on uploaded files and data sets, the IBM® Cognos® Analytics version 11.0.x data must be upgraded.

The upgrade process includes the following two tasks: retrieving some deeper data characteristics from data servers, packages, uploaded files, and data sets, and upgrading the Parquet file format in uploaded files and data sets.

Retrieve deeper data characteristics from data servers, packages, uploaded files, and data sets

The deeper data characteristics support the product functions that are behind the optimized user experience in dashboards, explorations, and other components. These characteristics are captured from samplings of data from the underlying sources.

Cognos Analytics 11.1 captures the deeper data characteristics for the following reasons:

To intelligently set the default column properties, such as Usage and Aggregate.
To provide recommendations for visualizations in dashboards, stories, and explorations.
To determine the subset of fields that are the best candidates to show in the relationship diagram in Explore.
To enable Assistant to be more successful in understanding the user's intent.
To provide other forms of automated assistance.

To retrieve the deeper data characteristics, you need to re-upload the Cognos Analytics 11.0.x sources using the following methods:

For data server connections, reload the schemas metadata.
Use the Load options option. Ensure that the following check boxes are selected: Retrieve the primary and foreign keys, Retrieve sample data, Retrieve statistics (version 11.1.4 and earlier).

For more information, see Preloading metadata from a data server connection.
For packages, use the Enrich package action.
Use the automatic enrichment option, and ensure that the check boxes Retrieve sample data and Retrieve statistics (version 11.1.4 and earlier) are selected on the Load options tab.

For more information, see Enriching packages.
For uploaded files and data sets, either run the ParquetUpgrade utility with option m or refresh the individual files and data sets manually.
The ParquetUpgrade utility with option m retrieves the deeper data characteristics from all uploaded files and data sets in the content store. When running this utility, you will upgrade the Parquet format in the affected files and data sets at the same time. For more information, see Running the ParquetMigrate utility.

For individual uploaded files, use the Append file and Replace file options. For individual data sets, use the Refresh option.

Upgrade the Parquet format in uploaded files and data sets

The Parquet file format that is used to store uploaded files and data sets has changed in Cognos Analytics version 11.1. The new Parquet format enables faster query processing on uploaded files and data sets.

You can implement this upgrade in the following ways:

Use the ParquetUpgrade utility to upgrade the Parquet format in all uploaded files and data sets in the content store.
Run this utility before users start running the reports, dashboards, or explorations. This ensures that all workloads immediately benefit from the performance gains associated with the new format. For more information, see Running the ParquetMigrate utility.
Manually refresh data in the individual uploaded files and data sets.
Use the Append file and Replace file options on uploaded files. Use the Refresh option on data sets.
Do not upgrade at all.
When a query uses data that wasn’t upgraded, the query service internally initiates the upgrade, and the users experience a one-time performance degradation when they run the dashboards, stories, reports, or explorations in Cognos Analytics 11.1. Subsequent queries use the upgraded data.

The new Parquet format is used automatically when new files are uploaded, new data sets are created, and when deployment archives that contain uploaded files and data sets are imported.