Databases Reference
In-Depth Information
Error Configurations for Processing
One of the challenges in designing a data warehouse is creating a perfect
schema without any referential integrity issues. However, this is often not pos-
sible and a significant amount of the time spent in designing a data ware-
house is typically spent in data cleansing. Analysis Services, by default, will
stop processing dimensions whenever it encounters specific referential integ-
rity issues. Some data warehouse designers might want to ignore the referen-
tial integrity issues by ignoring the records causing errors and include corres-
ponding fact data to Unknown member of dimensions so that they can see
the results of their cube design. Analysis Services gives you fine-grain control
for various referential integrity issues that can happen during processing. The
dimension property ErrorConfiguration allows you the fine grain control for di-
mension processing. If you click the ErrorConfiguration property and select
Custom, you will see all the properties that allow you fine-grain control as
shown in
Figure 8-21
.
Figure 8-21
The possible errors Analysis Services can encounter while processing a di-
mension are related to key attributes of the dimension. Typically, when you
have a snowflake dimension you can encounter dimension key errors while
processing whenever Analysis Services is unable to find corresponding keys
in the dimension tables involved in the snowflake schema. The main errors
that Analysis Services encounters are duplicate key errors (multiple occur-
rences of the key attribute in the dimension table), key not found error (unable
to find a key in the dimension table in the snowflake schema), and null keys
being encountered when you do not expect null keys to be present in the di-
mension tables. You can set properties to stop processing after a specific
number of errors have been reached, continue processing by reporting the er-
Search WWH ::
Custom Search