Data Stores on Separate Databases within DW.
2 posters
Page 1 of 1
Data Stores on Separate Databases within DW.
Is there an argument to have separate databases within a data warehouse for the following data stores:
Conversely, is there an argument to mix your operational data store with your dimensional data store for example? So, you would end up with a database with a mix of tables you would normally expect to find in a ODS with fact and dimension tables, which you would expect to find in a DDS?
Our DW is hosted by a supplier, and they are ooh-ing and ahh-ing over allowing us to create databases on the SQL server instance. Why, justification, etc. I would expect them to be on a separate database, not least for it to be more organised, and therefore, easier to support and maintain for example.
- Staging
- Normalised Data Store
- Operational Data Store
- Dimensional Data Store
Conversely, is there an argument to mix your operational data store with your dimensional data store for example? So, you would end up with a database with a mix of tables you would normally expect to find in a ODS with fact and dimension tables, which you would expect to find in a DDS?
Our DW is hosted by a supplier, and they are ooh-ing and ahh-ing over allowing us to create databases on the SQL server instance. Why, justification, etc. I would expect them to be on a separate database, not least for it to be more organised, and therefore, easier to support and maintain for example.
MarkW- Posts : 14
Join date : 2015-11-25
Re: Data Stores on Separate Databases within DW.
Are you talking about multiple databases on the same server or multiple servers?
Same server is just a logical subdivision that should not impact interoperability between the databases. If you are considering different servers, it is a matter of how much interaction occurs between the disparate systems. Generally speaking, you want staging to occur on the same server as the targeted data store. You would want to avoid joining tables across servers at all costs.
Same server is just a logical subdivision that should not impact interoperability between the databases. If you are considering different servers, it is a matter of how much interaction occurs between the disparate systems. Generally speaking, you want staging to occur on the same server as the targeted data store. You would want to avoid joining tables across servers at all costs.
Similar topics
» Data Warehouse spread across 4 separate databases
» Stores With Zero Sales
» Mirrored databases
» KDD (Knowledge discovery in databases)
» Large Fact Table and Maintaining Periodic Snapshot: Practice
» Stores With Zero Sales
» Mirrored databases
» KDD (Knowledge discovery in databases)
» Large Fact Table and Maintaining Periodic Snapshot: Practice
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum