Restrictions for Federated Partitions

Some functionality is not supported for Essbase cubes with a federated partition.

The cube must be within its own uniquely-named application. Federated partition cubes should not share an application with other cubes. Do not use the same Autonomous Data Warehouse schema for multiple instances of Essbase.
Before performing a data load through Essbase to the fact table, Oracle recommends you upload the data file to the Essbase Server. Client side data load is supported, but takes longer.
If you do not need to load data through Essbase to Autonomous Data Warehouse, you can use Data Tools in Autonomous Database to load data to the fact table and perform other management tasks. However, ensure that the cube outline and fact table do not get out of sync – see Metadata Precautions for Federated Partition Cubes.
Loading Essbase-formatted data export files into federated partition cubes can be time consuming. To optimize data loads, use a DBMS formatted source file. You can make one using the DATAEXPORT calculation command with DataExportCSVFormat option. CSV formatted files can be loaded faster because they are in accordance with DBMS_CLOUD package format options for source files.
The pivot dimension used in data load input files must be the same as the pivot dimension of the fact table.

See Federated Partition Data Load.
Importing data from multiple files in parallel using a MaxL import statement with wildcard characters is not supported for federated partition cubes.
Exporting a federated partition cube to an application workbook is not supported (does not export the data nor the partition definition).
Lifecycle Management (LCM) import operations (and Migration Utility import) are not supported for migration of federated partitions. Federated partitions must be recreated manually on the target.
Block calculation mode (enabled when Essbase configuration setting CALCMODE is set to BLOCK) is not applicable for federated partition cubes. Calculation processing is pushed to Autonomous Data Warehouse. If an exception exists and the calculation is processed on the Essbase Server instead, then solve order determines the dependency analysis.
When performing custom allocations on an aggregate storage cube with a federated partition, you can only override existing values. You cannot add to, nor subtract from, existing values.
Aggregate storage custom calculations and allocations are supported for federated partition cubes using only the MDX Insert logic. All restrictions documented for MDX Insert also apply to custom calculations and allocations in a federated partition cube.
Aggregate storage incremental data loads using buffers are not supported in a federated partition cube.
Block storage cubes must be in hybrid mode to support federated partitions. Do not configure ASODYNAMICAGGINBSO to any setting other than FULL for the application containing the federated partition, or else query results may be incorrect, and a warning message will be written to the log.
If you need to run Essbase block storage (BSO) calculation scripts, select a dense dimension as the pivot dimension. Calculation scripts are not supported for federated partitions if the pivot dimension is sparse.
For aggregate storage cubes, dimensions containing multi-level stored member hierarchies should not be selected as the pivot dimension. Select a pivot dimension with dynamic hierarchies, or a stored hierarchy that is a flat, single-level hierarchy (where all members are level 0 stored members).
Oracle Database has a limit of 1,000 columns, and the pivot dimension inherits this limit. Determine the number of eligible column members in the pivot dimension to ensure that you do not encounter the limit. The number of potential stored member combinations in the pivot dimension plus the number of dimensions in the cube should be less than or equal to 1,000.
The following calculation commands are not supported for federated partition cubes, and return an error if used:
- CALC AVERAGE
- CALC FIRST
- CALC LAST
- CCONV
- DATAEXPORTCOND
- DATAIMPORTBIN
- SET AGGMISSG OFF (Essbase always consolidates #MISSING for federated partition cubes)
- SET CLEARUPDATESTATUS
- SET CREATEBLOCKONEQ OFF (Essbase calculation of sparse dimensions is always top-down for hybrid and federated partition cubes, resulting in calculation of upper-level parents. In other words, the default behavior is SET CREATEBLOCKONEQ ON for federated partition cubes as well as hybrid cubes.)
- SET FRMLRTDYNAMIC
- SET REMOTECALC
- SET UPTOLOCAL
- SET UPDATECALC ON (Intelligent calculation, with its markers for dirty/clean blocks, is applicable only to non-federated, block storage cubes)
- THREADPARVAR
For more about calculation support, see Calculate and Query Federated Partition Cubes.
Calculation scripts using the @MDALLOCATE function are not supported and will fail with an error message.
Some long running calculations using IF/ELSEIF/ELSE logic may fail on federated partition cubes, returning either or both of the following ORA error(s) from Autonomous Data Warehouse:
```
ORA-04036: PGA memory used by the instance or PDB exceeds PGA_AGGREGATE_LIMIT
```
```
ORA-12805: parallel query server died unexpectedly
```
If you encounter such errors, you may need to increase Autonomous Data Warehouse hardware configuration to 16 CPU cores and 128G RAM. Refer to OCI documentation: Changing the Shape of an Instance.
Scenario management is not supported.
Transparent or replicated partitions against the federated partition cube are not applicable/not supported.
MaxL does not support creating or altering federated partitions, but you can use REST API.
MaxL statements and APIs for clearing/resetting data, clearing data regions, or clearing aggregates are not supported.
Text lists (a.k.a smartlists) are not supported
Request termination is not supported.
Varying attributes, and any default attribute calculation other than Sum are not supported.
MDX Sub Select is not supported.
Building aggregate views (MaxL statements execute aggregate process|build|selection) is not supported.
Merging data regions/slices is not applicable (because data is in Autonomous Data Warehouse).
Information returned from the MaxL statement query application APP-NAME list aggregate_storage storage_info (or equivalent API) is not complete/accurate.
Currency cubes are not supported.
Data audit trail is not supported.
Triggers on cube events are not supported.
Asymmetric queries may have slower performance.
Writeback performance (for example, the speed of submitting data updates from Smart View) can be slow if there is a large amount of data to submit.
Copying or renaming federated partition applications and cubes is not supported.
The following Essbase application or server configuration settings are ignored:
- AUTOMERGE
- AUTOMERGEMAXSLICENUMBER
- DATACACHESIZE
- CALCCACHE
- CALCCACHEDEFAULT
- CALCCACHEHIGH
- CALCCACHELOW
- CALCLOCKBLOCK
- CALCMODE
- CALCNOTICE
- CALCOPTFRMLBOTTOMUP
- CALCREUSEDYNCALCBLOCKS
- CALCPARALLEL
- CALCTASKDIMS
- DATACACHESIZE
- DYNCALCCACHEBLKRELEASE
- DYNCALCCACHEBLKTIMEOUT
- DYNCALCCACHECOMPRBLKBUFSIZE
- DYNCALCCACHEMAXSIZE
- DYNCALCCACHEONLY
- DYNCALCCACHEWAITFORBLK
- ENABLE_DIAG_TRANSPARENT_PARTITION
- EXPORTTHREADS
- FORCEGRIDEXPANSION
- GRIDEXPANSION
- GRIDEXPANSIONMESSAGES
- INDEXCACHESIZE
- INPLACEDATAWRITE
- PARCALCMULTIPLEBITMAPMEMOPT
- SSAUDIT
- SSAUDITR
- SSLOGUNKNOWN
- SUPNA
- TARGETASOOPT
- TARGETTIMESERIESOPT
Creating a federated partition may fail with the following error if too many levels exist in the Essbase outline: Remote warning from federated partition on Analytic View: [ORA-04063: hierarchy has errors].
Federated partition creation may fail if characters or name lengths used in Essbase dimension names or member names in the pivot dimension are not supported or are considered special by Autonomous Data Warehouse. These limitations should be considered in addition to the documented Essbase Naming Conventions for Dimensions, Members, and Aliases.
You can only delete a federated partition from the Essbase web interface. You cannot delete it from Cube Designer.