Hi Ashwath ,"The Snowflake model has more joins between the dimension table and the fact table, so theperformance is slower. ... The Star model, on the other hand, has fewer joins between dimension tables and the facts table."But when you are going to pull the data in tableau you will use Extract and Extracts are flat file .Now we have hyper as well .So for the first time it may take some time after that tableau is going to work on that extract .Also sharing few discussion that will clarify your doubt ::-Nikher
You don't HAVE to use an extract, Tableau can do a live connection to both star & snowflake and if you define the join in the data source pane then Tableau will know how to fire the correct SQL at the database.
Personally, I like to flatten the schema in a view and then you don't have to join in Tableau ... this is a LOT easier for users that aren't that techy anyway.
Hi Chris ,
Wanted to ask one question on live and extract connection ,as per my experience if we are pulling data or using extract we can do the things in more optimized way as the performance is not dependent on database .
So in case we are using star or snowflake,Please clarify when to use live connection or when to use extract If I don't want my performance to be dependent on DB .
Thank you .
Definitely if you're using a live connection Tableau can only be as fast as the database. In a lot of circumstances when the database is slow (network latency or a number of other factors) then changing to extracts is better BUT it depends on the data. If you want "real time" dashboards then extracts might take too long, if you don't want performance to be dependent on the db then use extracts