Databases Reference
In-Depth Information
Dimensions can be incrementally processed using the ProcessIncre-
mental or ProcessAdd options. ProcessIncremental retrieves the data
from the relational data source, compares the data to existing dimension
members on Analysis Services, and then makes updates to existing di-
mension members if there are any changes. If there are new mem-
bers, they are added to the dimension and this does not affect the par-
titions. However, if the dimension members are updated such that the
relationship between attributes have changed (an employee's marital
status changed from single to married) then the aggregations for the
corresponding partitions will be dropped. The cube will still be available
for querying but it can impact the performance of the queries which
were using the dropped aggregations. ProcessIncremental takes more
time than Full process of the same dimension because Analysis Services
does the additional work of checking for updates for existing members;
however, it provides you the flexibility of having the dimension available
for querying and the cube does not have to be reprocessed. ProcessAdd
is a new option in Analysis Services 2005 for processing dimensions.
This option allows you to add new dimension members that have been
added in the relational data source to existing processed dimensions.
We recommend using the ProcessAdd option for dimensions whenever
you have new members being added to the corresponding dimension
tables in the relational data source. ProcessAdd for dimensions is useful
in cases where new products are added to the products table on a peri-
odic basis. You need to specify relational queries to the data source that
will return the new rows that have been added since the last dimension
update. The query to retrieve new data along with the data source and
data source view elements is to be specified in the DDL along with the
ProcessAdd called as out of line binding. Enclosed below is an example
of using DDL using ProcessAdd with out of line binding.
<Batch xmlns="http://schemas.microsoft.com/
analysisservices/2003/engine">
<Parallel>
<Process xmlns="http://schemas.microsoft.com/
analysisservices/2003/engine">
<Object>
<DatabaseID>Adventure Works DW</DatabaseID>
<DimensionID>Dim Customer</DimensionID>
</Object>
Search WWH ::




Custom Search