IBM Support

Enhancements in Cognos BI 10.2.1 Fix Pack 3

Product Lifecycle


IBM Cognos Business Intelligence Fix Pack 3 (FP3) enables many valuable enhancements related to the dynamic query mode (DQM) including Dynamic Cubes.

An overview of these enhancements is presented in this document.


Support for new JDBC data sources

  • Cloudera Impala
  • DB2 iSeries JT400 JDBC driver (IBM Toolbox for Java)
  • Actian ParAccel
  • EXASOL EXAsolution
  • HP Vertica
  • Pivotal Greenplum

Setting up connectivity for data sources using the JDBC connection type
Refer to the Software Compatibility Report (SPCR) for Cognos BI 10.2.1 for updated details of all supported environments.


Publish TM1, Essbase and Analysis Services packages from Cognos Administration
Prior to FP3, publishing a package for a IBM Cognos TM1, Microsoft Analysis Services, Oracle Essbase data source necessitated going through an import exercise in Framework Manager, creating a package and then publishing. Now you have the ability to simply create a data source in Cognos Connection and then automatically create a package that can be used for reporting such that there is no need to use Framework Manager to perform the import/publish operation.


Filler members for TM1 cubes with element-level security
If you are using IBM Cognos TM1 server at version 10.2 FP1 ( or higher, now you can allow "filler members" to be generated to fill gaps due to restricted access from the root of the hierarchy down to those members whose data should be visible to the user. This enables a TM1 cube with element-level security to be more accessible for self-service analysis in Cognos BI.

See technote 1660287 for more details.


Filter join optimization for Report Studio

The purpose of the filter join optimization is to improve performance of joins by filtering one side of the join using the values retrieved by the other side. This can help you minimize the amount of data retrieved by the Cognos BI server, particularly when querying heterogenous data sources where one side of the join has orders of magnitude less rows than the other side. Applying filter joins in such cases allowed reports to run between 10 and 100 times faster in IBM Lab testing.

See technote 1660292 for more details.


Support for user defined aggregate functions
In all versions of Cognos BI, if an unrecognized function is encountered when the BI server parses an expression in a report or model, it will submit that function to the database in good faith. If it is a valid scalar function on the database, it is processed successfully and the desired results are displayed in the report. Previously, only scalar functions were supported because the BI server was unable to distinguish between scalar and aggregate functions. With 10.2.1 Fix Pack 3, now you can also call upon aggregate functions in your database by using using the keyword aggregate followed by a colon before the function name.

See technote 1660309 for more details.


Near real time updates for Dynamic Cubes
New data records added to a fact table can now be applied incrementally to a Dynamic Cube on demand. The in-memory aggregate cache and the data cache of a dynamic cube are updated without requiring the restart or refresh of the cube, or updates to the database summary tables. See techdoc 7040299 for more details.
A new command line tool, DCAdmin, can be used to perform all administrative commands against a dynamic or virtual cube, including the ability to apply a specific set of fact data updates to a cube. Cube metrics can also be obtained through this tool. See techdoc 7040451 for general details on the DCAdmin tool, and techdoc 7040417 on loading incremental updates.
A cube can be paused (leaving all data caches intact) while database summary tables are updated to align with the fact table updates. Once updates are completed, the cube can resume operation. See techdoc 7040535 for more details.

See techdoc 7040458 for guidance on maintaining aggregate tables in your database when incrementally updating your dynamic cube.


Relative time functionality for Dynamic Cubes
New relative time functionality aligned with that available in IBM Cognos Transformer, adding support for:
      • Members that correspond to future time periods.
      • Single period custom members e.g.” Same Month, Prior Year”
      • Period-to-date custom members e.g. “Quarter to Date, Prior Year”
      • N-period running total members e.g. 6 month rolling window
As well, you now have the ability to control which relative time members are added to a time hierarchy.

See techdoc 7040513 for details on predefined relative time members
See techdoc 7040594 for details on custom relative time members.
See techdoc 7040479 for details on new time level types: Holiday, Season, Semester, Trimester.

Named sets for Dynamic Cubes
Named sets can now be defined in Cube Designer in the context of a Dynamic Cube or Virtual Cube. Named set expressions can be any valid member set expression and can also make use of macro expressions.

Named sets are accessible in any of the authoring interfaces of Cognos BI including Report Studio and Cognos Workspace Advanced.

See techdoc 7041007 for more details.


Memory utilization improvements for Dynamic Cubes
The size of the member cache has been reduced, requiring approximately 550 bytes per member.

Dimension members can now be optionally shared between dynamic cubes and virtual cubes that reside on the same server to reduce the overall memory footprint. See techdoc 7040198 for details on shared dimensions.

The query service now recognizes when the JVM heap is nearly full and will attempt to release resources from in-memory caches in an attempt to avoid the occurrence of out of memory exceptions.


Aggregate Advisor improvements for Dynamic Cubes
It is now possible to merge multiple in-memory recommendations into a single recommendation from the Aggregate Advisor that is published to the content store.

Aggregate Advisor recommendations are now stored within the query service as opposed to within the Dynamic Query Analyzer. One benefit of this is that if the communication channel between DQA and the Cognos BI server is lost, the advisor can still complete its work.

See technote 1659240 for more details.


Create, edit, and publish packages in Cube Designer
The new Packages object in Project Explorer contains model packages. To add or remove cubes to or from a package, use Project Explorer or the package editor. You can edit a package's name, comments, and location. To publish a package, right-click it, then click Publish.


Model virtual cubes through the Dynamic Cubes API
You can now model virtual cubes using the REST based API for Cube Designer.

See technote 1659500 for more details

Related Information

[{"Product":{"code":"SSEP7J","label":"Cognos Business Intelligence"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Component":"--","Platform":[{"code":"PF002","label":"AIX"},{"code":"PF010","label":"HP-UX"},{"code":"PF016","label":"Linux"},{"code":"PF027","label":"Solaris"},{"code":"PF033","label":"Windows"}],"Version":"10.2.1","Edition":"","Line of Business":{"code":"LOB10","label":"Data and AI"}}]

Document Information

Modified date:
15 June 2018