Tabular model – Workspace and Deployment server

In Visual studio when working with a new tabular project you get a local sandbox analysis cube database to play with your dimension, hierarkies and measures. You can also choose to work locally or remote for your workspace. If you choose to work locally, you can continue to work when you are offline. But if the project is heavy and you need lots of data in the workspace model to test formulas, you might want to have the workspace remote on a “bigger” analysis server installation. Bevare of losing the workspace db though. Took me some time to figure out how to get on track again after “laborating” with loss of AS db for the workspace. But of course you have backups of your cube databases? No!? Well, you should! 🙂

The trick was to create a new project and use the already deployed version to create the model.bim file. which in turn gave me another cube name and a aligned new workspace database, and it just worked after refreshing all data connections. The .bim file contains the model, and everything is created out of that one, so be careful with it. So – how do I find where my default workspace server and deployment server is configured? * default workspace server: Go to “Tools”-menu, “Options”, “Analysis Services”, “Data Modeling”, “Default workspace server” There you also can chose retension strategy for what will happen after you close the model. Another way to reach the configurations for the project are via the project menu: * analysis server name for deployment Go to: “Project”-menu, “<cube name> Properties”, “Deployment Server”, “Server” Visual Studio is a very rich environment to work in. As a SQL DBA I am not that familiar with it, but I begin to get the feel of it and I like it!

Good luck with your models and feel free to contact one of our BI/SQLServer consultants if you need help.

/Jonas Bergström