A deploy process generates a log of all the steps involved. However, there are sometimes request for more detail about what kind of updates and queries are taking place in the sql database during the steps of a deploy process. Following is more detail behind some of the deploy steps.
Build dimension for each dimension
- Update DS and HS tables from S_DS and S_HS tables.
For each hierarchy, read HS table and build HL table entries in memory.
Recreate HL table based on levels in hierarchy and write records.
Update DimensionHierarchies and DimensionHierarchyLevels tables.
Create records for HC table.
Update FACT table for each model
- Create FACT table if missing
If dimensions added to or removed from the model add or remove fields in the FACT table.
If index missing or if fields added to or removed from the FACT table, rebuild the index.
Create View of FACT table
Processing of AS database
- Dimensions – Analysis Services generates SQL queries to read members, properties, and hierarchies for each dimension.
Models – Analysis Services generates a SQL query for each partition of each model to read data from the FACT table to process the model. Each query is displayed in the deploy log twice. For large applications with large amounts of data, you can control the size of each of these queries by defining Model partitions.
- Update SecrityRoleMembers and Users tables.
Update SecurityFactTable table.