Why Heterogeneous?

View previous topic View next topic Go down

Why Heterogeneous?

Post  vueman00 on Fri Oct 15, 2010 1:00 pm

Hi,

I am a fresh student and I am trying to understand the Heterogeneous concept. I am wondering what the difference is between a Heterogenous schema and a regular schema. Why would someone use a heterogeneous vs a regular schema and when is it appropriate to use one or the other?

Thanks much for your input.

Vue

vueman00

Posts : 3
Join date : 2010-10-15

View user profile

Back to top Go down

Re: Why Heterogeneous?

Post  BoxesAndLines on Fri Oct 15, 2010 1:39 pm

Heterogenous normally refers to different sources for the same class of information. For example, most companies have multiple channels for managing customers. Each of those siloed applications will have their own interpretation of the data model and underlying data structures. In the EDW, I need to reconcile the heterogenous sources to present a consistent, unified, view of the customer.
avatar
BoxesAndLines

Posts : 1212
Join date : 2009-02-03
Location : USA

View user profile

Back to top Go down

Re: Why Heterogeneous?

Post  vueman00 on Fri Oct 15, 2010 1:46 pm

would this be the same as Snowflaking?

vueman00

Posts : 3
Join date : 2010-10-15

View user profile

Back to top Go down

Re: Why Heterogeneous?

Post  ngalemmo on Fri Oct 15, 2010 2:27 pm

No. Snowflaking is a form of star schema where a dimension table contains a foreign key to another dimension table. That is all the term means.

I've never used the term 'heterogeneous' when describing a data warehouse schema... it would seem to me such an approach (i.e. a heterogeneous data warehouse) would be counter-productive, as one of the goals of DW design is to integrate heterogeneous sources, rather than keep them that way.

avatar
ngalemmo

Posts : 3000
Join date : 2009-05-15
Location : Los Angeles

View user profile http://aginity.com

Back to top Go down

Re: Why Heterogeneous?

Post  ngalemmo on Fri Oct 15, 2010 2:44 pm

To carry this discussion a bit further, take an example of a company that has separate product lines within separate divisions with separate sales systems (orders, product master, customer master, inventory, etc...).

These sources are heterogeneous... different data, different data structures and so forth, but they both cover 'sales' from a corporate point of view. In a well designed data warehouse, these sources would be integrated such that data can be viewed and analyzed seamlessly across divisions using standard as well as divisional interpretations, yet at the same time, retain the identity and nomenclature of the source systems.

My guess is, a heterogeneous data warehouse model would keep everything separate, not attempt to conform data interpretation and leave it up to the users to figure out how to combine the data. Not a very productive environment...
avatar
ngalemmo

Posts : 3000
Join date : 2009-05-15
Location : Los Angeles

View user profile http://aginity.com

Back to top Go down

Re: Why Heterogeneous?

Post  vueman00 on Fri Oct 15, 2010 4:10 pm

Thanks for the clarification. I appreciate it!

vueman00

Posts : 3
Join date : 2010-10-15

View user profile

Back to top Go down

Re: Why Heterogeneous?

Post  Sponsored content


Sponsored content


Back to top Go down

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum