locked
Generate change script for the cube RRS feed

 > 

  • Question

  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi gurus

    I have a cube and deployed to Dev, Test and Live. Now, I have to add new attribute to an existing dimension. I did in SSDT and deployed to Dev.

    I need to deploy this change to Test and Live. How can I generate change script for the cube so that I can deploy only the changes and process only the changed dimension instead of re processing whole cube?

    I have tried Deployment wizard but I need to re process whole cube.

    How can I do this?

    Thanks in advance

    Wednesday, January 20, 2016 1:27 PM

Answers

  • Question
    Sign in to vote
    1
    Sign in to vote

    When you create a script and use the Default processing option, ssas will only process your changed dimension and all the measure groups connected to it. This might seem like the entire cube, or effectively be the whole cube.

    Unfortunately, there is no way around reprocessing the dimension and related measure groups if you have changed the structure of the dimension.

    By the way, if you just rename attributes, add translation, add calculations etc, the Default processing option will not involve reprocessing the measure groups and dimensions.

    Hope that helps,

    PS, you might find this blog useful in minimising downtime during cube deploys.

    http://richardlees.blogspot.com.au/2009/05/new-analysis-services-database-version.html

    Richard


    Thursday, January 21, 2016 3:29 AM
  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi Rajkm,

    In this scenario, if this added attribute is not used other objects like measures, you can just process and deploy the dimension only. Otherwise you have to deploy all dependent objects. In SSDT, you can directly set "Deploy changes only" under Deployment setting. And you can use Process Default, it will process changed objects only with Process Full operation.

    Regards,

    Simon Hou
    TechNet Community Support


    Thursday, January 21, 2016 6:03 AM

All replies

  • Question
    Sign in to vote
    1
    Sign in to vote

    When you create a script and use the Default processing option, ssas will only process your changed dimension and all the measure groups connected to it. This might seem like the entire cube, or effectively be the whole cube.

    Unfortunately, there is no way around reprocessing the dimension and related measure groups if you have changed the structure of the dimension.

    By the way, if you just rename attributes, add translation, add calculations etc, the Default processing option will not involve reprocessing the measure groups and dimensions.

    Hope that helps,

    PS, you might find this blog useful in minimising downtime during cube deploys.

    http://richardlees.blogspot.com.au/2009/05/new-analysis-services-database-version.html

    Richard


    Thursday, January 21, 2016 3:29 AM
  • Question
    Sign in to vote
    0
    Sign in to vote

    Hi Rajkm,

    In this scenario, if this added attribute is not used other objects like measures, you can just process and deploy the dimension only. Otherwise you have to deploy all dependent objects. In SSDT, you can directly set "Deploy changes only" under Deployment setting. And you can use Process Default, it will process changed objects only with Process Full operation.

    Regards,

    Simon Hou
    TechNet Community Support


    Thursday, January 21, 2016 6:03 AM