Update-driven approach is chosen
4 posters
Page 1 of 1
Update-driven approach is chosen
a. Update-driven approach is chosen rather than query driven approach while integrating multiple heterogeneous information sources.justify whether true or false ?
b. Describe benefits and drawbacks of a source-driven architecture for gathering of data at a data warehouse , as compared to a destination-driven architecture ..
b. Describe benefits and drawbacks of a source-driven architecture for gathering of data at a data warehouse , as compared to a destination-driven architecture ..
aashaya m- Posts : 18
Join date : 2011-04-25
Yes sir......
Yes sir...in our Elective we do have Data warehousing ..but little bit confused about these questions ..you have any hint please reply
aashaya m- Posts : 18
Join date : 2011-04-25
Re: Update-driven approach is chosen
My guess is question A is asking which is better: handling integration at query time or when loading the DW.
Also not completely sure about B but I suspect the question is about what drives the design: the data available from the sources or the stated requirements of the mart.
It would be easier to understand the questions if I heard the lecture...
Also not completely sure about B but I suspect the question is about what drives the design: the data available from the sources or the stated requirements of the mart.
It would be easier to understand the questions if I heard the lecture...
Re: Update-driven approach is chosen
Nice questions. At least folks are starting to think about these types of things.
BoxesAndLines- Posts : 1212
Join date : 2009-02-03
Location : USA
Re: Update-driven approach is chosen
B sounds like comparing the pros and cons of Star Schema Vs 3nf. Leaving data in tables based on the source system rather thaa integrating the data from the various source systems into the same tables.
One is easier to build but harder to use. The other is easier to use but harder to build. One gets data quicker to the end user but forces the user to do a lot more stuff to the data in order to use it. The other takes time to build, harder to get real time data to user, but is a lot easier for the user to actually put into use.
One is easier to build but harder to use. The other is easier to use but harder to build. One gets data quicker to the end user but forces the user to do a lot more stuff to the data in order to use it. The other takes time to build, harder to get real time data to user, but is a lot easier for the user to actually put into use.
Jeff Smith- Posts : 471
Join date : 2009-02-03
Re: Update-driven approach is chosen
Agree. Replicating 3nf models is just a easy way to push the problem onto another group, in this case the users and/or reporting team. Its an expensive way to save a few bucks up front.
Re: Update-driven approach is chosen
I am not sure it's about saving a few bucks. I think it's about implementing the fastest and easiest solution. End User pain doesn't fit into the traditional IT equation. I see it all the time. A report that takes 3 minutes to run is preferable over a report that needs a new aggregate table to run in 5 seconds.
Jeff Smith- Posts : 471
Join date : 2009-02-03
Similar topics
» Data Vault v's Dimensional Model
» What to do when facts change? Update the fact table with update and deletes? or use SCD type 2?
» metadata driven ETL architecture
» Modeling tool with support for hierarchies
» Automating dimension and fact loading ETL at the database level - Good or Bad?
» What to do when facts change? Update the fact table with update and deletes? or use SCD type 2?
» metadata driven ETL architecture
» Modeling tool with support for hierarchies
» Automating dimension and fact loading ETL at the database level - Good or Bad?
Page 1 of 1
Permissions in this forum:
You cannot reply to topics in this forum