Lately, I've been trying to clean up a group of projects that my company received from an outside contractor. Among them is an SSAS cube, with an associated data warehouse.

The contractors generated the cube top-down, using the Schema Generation Wizard to generate the schema for the data warehouse from the cube. Normally, based on what I've read about the SSAS tools in Visual Studio, you make changes to the data warehouse by changing the cube, and then re-running the Schema Generation Wizard to update the data warehouse. I'd like to maintain the database using the SSAS tools so that we have one centralized place to manage both the cube and data warehouse's schema.

My goal with cleaning up these projects is to get them to run on developers' local machines. That will require standing up a full new set of independent databases on each developer's machine. I want to manage the data warehouse schema from the SSAS tools, so I'd like to fully regenerate the schema each time we stand up a new developer.

I'm not concerned with data or stored procedures in the data warehouse at the moment. I'd be fine with completely deleting the data warehouse and regenerating it when needed. Deleting the .dsv file will not be an option, however, because it contains queries against the table structure defined by the cube.

Is there any way to regenerate a previously-generated schema using the Schema Generation Wizard, without deleting the .dsv?

  • 2
    How is this question different than dba.stackexchange.com/q/204087/79569? My answer there still is how I would answer here. – GregGalloway May 2 '18 at 1:10
  • 3
    I didn't get a satisfactory answer, so I changed that question and added a bounty. A moderator, Paul White, rolled back the question and the bounty and told me to ask a refined version of it. I'm really looking for the "how", moreso than whether this is a good approach or not. – Julian Goldsmith May 2 '18 at 13:33
  • @TomV I explained how I would recommend doing it on the other thread. Apparently julian-goldsmith wants to do it another way. – GregGalloway May 2 '18 at 18:21
  • @greggalloway: I was looking for how to use the Schema Generation Wizard to do this. I have other ways I can do it, but I'm trying to manage the cube entirely from one place. – Julian Goldsmith May 2 '18 at 18:42
  • 1
    I would do that as a one-off. Take a copy of the project locally then create the db schema. Next import it into a Visual Studio Database Project. Include that in the same solution as the cube if you like. Now maintain the database schema using the database project and the cube via the multi-dimensional project. I think you might be creating an artificial barrier for yourself by saying you only want to use the "Schema Generation Wizard" tooling because, as you've pointed out, doing it over and over again is not that straightforward. Make sense? – wBob May 2 '18 at 20:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.