February 14, 2011 at 2:28 pm
Hello all,
I am bit confused... I know we develop OLAP using Fact and Dimension tables. Do we call the database that contains Facts and dimension tables as a datawarehouse? (For eg: Do we call AdventureWorksDW database as datawarehouse?)
Thanks.
February 23, 2011 at 12:28 pm
Yes the data warehouse is often considered the database. However, the surrounding systems are often referred to by end users as the data warehouse.
--------------------------------------------------------------------------
When you realize you've dug yourself into a hole....Step 1...stop digging.
February 23, 2011 at 12:34 pm
Rookie R (2/14/2011)
I am bit confused... I know we develop OLAP using Fact and Dimension tables. Do we call the database that contains Facts and dimension tables as a datawarehouse? (For eg: Do we call AdventureWorksDW database as datawarehouse?)
A datawarehouse is by definition a database designed to support reporting in an organized and well performing way.
There are basically two schools when coming to the design of a datawarehouse, those are the Kimball and the Inmon approaches to it.
Kimball approach is based on dimensional modeling - also called star modeling -and datamarts which usually include a FACTual table in the center of the star which is sourrounded by DIMensional tables.
If you see FACT and DIM tables you are most likelly looking at a Star schema database developed somehow around Kimball's approach to it.
Hope this helps.
_____________________________________
Pablo (Paul) Berzukov
Author of Understanding Database Administration available at Amazon and other bookstores.
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.February 23, 2011 at 11:05 pm
I would add, just because you have fact and dim tables does not make it a data warehouse. I've worked on warehouse before and the definition there is that it includes ALL the data of the company in one place. i.e. details of your clients, what they bought, your sales people, theirs sales, thier salaries etc. so really all the data from all your system in your company.
Currently I'm working in what I would rather call a managment information system. It is for 1 department; which brings in about 20% of the total company revenue, but it has a lot of clients (10 million) but it still not a data warehouse as I'm not accessing all of the company data floating around.
Personally, whether snow or star schema, fact or dim tables, I would only call it data warehouse when you have all the company data at your disposal.
February 24, 2011 at 8:15 am
From my perspective a Data Warehouse is what it says on the tin. A warehouse of data.
This normally means it a central repository of data from 1 or more (normally More) data sources collected together in one place to support a function.
MI is a primary function of a data warehouses but they are also great for any kind of read only access like Website etc.
I think the primary difference between a data warehouse and a Database is that the warehouse is a refreshable, read only image of 1 or more data sources data.
February 24, 2011 at 8:31 am
mark.marsh (2/24/2011)
I think the primary difference between a data warehouse and a Database is that the warehouse is a refreshable, read only image of 1 or more data sources data.
Very true, forgot to add that point.
February 24, 2011 at 8:34 am
chris.stuart (2/24/2011)
mark.marsh (2/24/2011)
I think the primary difference between a data warehouse and a Database is that the warehouse is a refreshable, read only image of 1 or more data sources data.Very true, forgot to add that point.
Read-Only ? , how do you load the data into the warehouse if it's read-only?
February 24, 2011 at 8:44 am
steveb. (2/24/2011)
chris.stuart (2/24/2011)
mark.marsh (2/24/2011)
I think the primary difference between a data warehouse and a Database is that the warehouse is a refreshable, read only image of 1 or more data sources data.Very true, forgot to add that point.
Read-Only ? , how do you load the data into the warehouse if it's read-only?
Read Only applies to the end users. As opposed to a OLTP database where users enter data (Process Transactions). Of course it's not totally read-only if you include the refreshing of data.
February 24, 2011 at 8:46 am
Read-Only ? , how do you load the data into the warehouse if it's read-only?
Not Read-Only as in Read-Only, Read-Only as in Read-Only. 😀
Read-Only as in data will never be updated on your warehouse from a front-end and then exported to another system.
Read-Only as in data will only be received from other transactional system, i.e. all updates, deletes and inserts happens on a transactional system, but never in your DB.
Read-only as in Read-only also a requirement..:hehe:
February 24, 2011 at 8:50 am
chris.stuart (2/24/2011)
Read-Only ? , how do you load the data into the warehouse if it's read-only?
Not Read-Only as in Read-Only, Read-Only as in Read-Only. 😀
Read-Only as in data will never be updated on your warehouse from a front-end and then exported to another system.
Read-Only as in data will only be received from other transactional system, i.e. all updates, deletes and inserts happens on a transactional system, but never in your DB.
Read-only as in Read-only also a requirement..:hehe:
all clear now.. you had me worried for a minute,,,
February 24, 2011 at 4:24 pm
Thank you all!!!
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply