I want to build a data warehouse solution based on Anchor Modeling.
But I have a lot of different sources. The data from different sources will be split over multiple staging tables. Sometimes the same Business Key can be found in multiple staging tables with different attributes.
Lets say it is not possible to combine the information from different sources into one staging table before I load it into AnchorModeling. One of the reasons could be that the combined amount of attributes would exceed 256 attributes which is a limitation of the views in Anchor modeling.
Is it not a solution to make it possible to add a (optional) "mnemonic group name" to each attribute (and take the last entered as a default). Attributes can then be grouped in views with this mnemonic group. For instance place the code at the end of the view name. This can be backward compatible but once you start adding the group mnemonic only then the view name will be different.
* You can use the Mnemonic groupname to generate ETL code for loading data from different sources
* Most important advantage. I can have > 256 attributes linked to one anchor as long if I do not exceed the 256 limit in one group.
Or is there a better way how to deal with the limitation of 256 attributes for one anchor?
Thanks in advance!