Overview of Load Jobs

Load Jobs is a server driven method to load data into the OLAP database. The server executes Load Jobs on a schedule. The job descriptions are stored in a relational database.

The information needed to import data is stored independently from OLAP. You can setup this database without being connected to OLAP. When the information is ready, its location is registered to OLAP. The jobs stored there are executed, either on demand or on a schedule.

The Load Jobs database does not have to be the same database as the database that is the source of the data. Usually, two separate databases are used. The Load Jobs database has small tables containing the information needed to load the data from the data source.

The dimensions must be created before the facts are loaded. Once the dimensions are in place, the facts can be loaded. This table describes the process:


Task	Job type	Description
Load dimension	1	For each one of the dimensions. A dimension can be a simple list or a complex structure with multiple hierarchies and attributes.
Create cube	2	A task that specifies which dimensions belong to a cube. The dimensions are created before the cube can be defined, but no facts are loaded initially.
Load facts	3	Import fact data into a cube whose dimensions have been created.

These tables describe the dimensions:

_Dimensions
_DimensionLevelNames
_AttributeFields
_Subsets
_Hierarchies

These tables are used for cubes and fact loading jobs:

_Cubes
_CubesDimensions
_CubeAccessControl
_FactLoad
_FactLoadParameter

The metadata to transfer are stored in these tables:

_Jobs
_ScheduledJobs
_Sources
_Dimensions
_DimensionLevelNames
_Cubes
_CubesDimensions
_AttributeFields
_Subsets
_CubeAccessControl
_JobsParameters
_FactLoad
_FactLoadParameters
_Hierarchies
_Parameters
_ErrorText
_ErrorLogAttributeFill
_ErrorLogFactLoad
_ErrorLogElementLoad
_ErrorLogOther

The source data in the relational database can be in a table or a view.