2 Replies Latest reply on Dec 20, 2013 12:10 PM by Murali Govindu

    Data source performance: Is there any performance difference between data sources local to a workbook vs. connecting to data sources published on the server?


      I am wondering or not there is a performance difference between having a workbook that is available on Tableau Server connect to data sources local to that workbook (with this image next to the data source:t-local.png) vs. connect to the same data sources that are published on the server (with this image next to the data source:t-published.png)?


      - I will not be using extracts as I need to connect live to my database (Redshift).

      - I am aware that calculated fields which span data sources will not be available when pulling a data source down from the server.  This is OK as I have a master workbook which I build most reports off of.


      I have noticed that when I publish the data sources to the server and re-open an old workbook that used the same data sources (local to the workbook but have since been updated) the workbook lets me know that there is a new version of the data sources on the server and asks if I would like to update the data sources in this workbook.  When I hit OK, any changed calculated fields (based off of fields coming from only one data source) get updated and any new fields are added.  This tells me that Tableau knows they are the same data source. 


      So, is there any performance differences or limitations of either approach I should know about?