Troubleshooting Processing (Analysis Server)
What problem are you having?
Changing a data source provider (for example, moving from an ODBC provider to a Microsoft® SQL Server™ provider) for an existing cube causes an error the next time the cube is processed.
Cause: Different providers sometimes use incompatible SQL dialects, and a statement that was valid for one provider is not valid for another.
Solution: The error message should identify the syntax that is not valid. Use Decision Support Objects (DSO) or Cube Editor to correct the problem. To avoid such problems, do not change the data source provider for an existing cube.
Cause: One or more members of a dimension contains more than the maximum number of children allowed per parent.
Solution: Create member groups to provide intermediate levels that organize the dimension such that no member has more than the maximum allowable children. For more information, see Creating Member Groups.
While processing a cube, you received one of the following error messages:
- If the dimension is not parent child dimension: "A member with key key was found in the fact table but was not found in the level level of the dimension dimension."
- If the dimension is a parent child dimension and the MembersWithData property of the dimension is set to dataforLeafMembersOnly, the message is: "A member with key key was found in the fact table but was not found in the set of leaf members of dimension dimension."
- If the dimension is a parent child dimension and the MembersWithData property of the dimension is not set to dataforLeafMembersOnly, the message is: "A member with key key was found in the fact table but was not found in the dimension dimension."
Cause: The dimension table does not contain a member related to a fact in the fact table, or the dimension was not processed after changing the dimension table.
Solution: Verify referential integrity between dimension tables and fact tables. Process a dimension after changing the dimension table and before processing cubes that use the dimension. For more information, see Optimizing Cube Schemas.
To manage a large number of these key errors, you can log the missing members to a file and use Data Transformation Services (DTS) to import the log into a relational database. You can then either remove the records with the keys from the fact table or add them to the dimension table. You select to log dimension key errors to a file in the Cube Processing Settings Dialog Box.
Cause: The source data lacks referential integrity. For example, the cube's fact table contains foreign key values that are not present in a joined dimension table's primary key column.
Solution: Correct the referential integrity of the source data so that inner joins between the fact and dimension tables include all rows in the fact table.
There are two possible causes.
Cause: The cube or virtual cube was changed but not processed.
Solution: First the administrator must process cube or virtual cube. After the cube is processed, the user must reconnect to the server computer.
Cause: The cube or virtual cube was processed while the user was browsing the cube.
Solution: The user might need to reconnect to the server computer. Reconnection is required for a full update but not for an incremental update.
Cause: There is a many-to-many instead of a one-to-many relationship between the dimension table and the fact table.
Solution: Remove redundant members from the dimension tables.