I develop some application that has database wery generic so really can't use it for reporting. So I need solution how to create开发者_StackOverflow reporting. I'm developer so my knowledge in DBA domain is bounded. For now I have ideo to create another database where I'll pu denormalized data from original db. So I saw that I could use SSIS for that and woul be glad if someone could give me some advice how to attack that problem. Should I sync data once a day and run reports that way. Is there solution to sync data allways so reports would be up to date? Please any advice.. Thanks!
Damir,
What I get from your message is that you are getting close to build a Datawarehouse using a Star Schema pattern.
You could have two databases, One with normalized data and the other one with the Star Schema pattern (Your DW), and then create a script that would use your normalized data and put them in your datawarehouse. For the frequency of your script it is up to you : After each transaction, every hour, once a day, etc...
The advantage of having a datawarehouse is that you will be able to use OLAP cubes and the MDX language for your reports. It's a plus !
Hope it could help,
If you are on sql server 2005 or greater, explore Merge statement. For smaller tables, just truncate and reload. 'Smaller' could be subjective - but if takes less than 2-3 minutes to load, that could be termed as small. Obviously, during that period any query that uses such tables would fail.
精彩评论