Q1. This should give you the mdf and ldf you need for the SQL database.
http://adventureworksdw2008.codeplex.com/
Q2. Yes, the datasource should be connecting to the SQL database that was just attached. If you right click the project and go to the properties, you should see an option for deployment where you can define the server and Analysis Services database name you want to the deploy this project as.
Q3. Importing is the way to get a copy of a deployed SSAS database. If you just wanted to look at the code, you can connect directly to an existing SSAS database. In BIDS, File > Open > Analysis Services Database (or something named similar) and it'll ask for the server and database you want to connect to. This will open the live SSAS database. Any changes you save will be applied directly to the database on the server.
Overview of SSAS (MOLAP). (Molap is the storage setting in which structure and aggregate data is stored in the SSAS databse.)
SSAS is essentially an aggregated view of underlying SQL data. The storage of data in SSAS when using MOLAP can be thought of as massive complex indexes. All the definitions you create in the SSAS database are just indexes. To process and update these indexes, you have connect the SSAS objects (Measures and Dimensions) to the underlying data tables from the SQL database. The SQL table DimDate has all the dates in all the different formats you would want to see it. In SSAS, the Date(Time) dimension references the DimDate table and defines which columns are used and the format. SSAS dimension also allows you to define a hierarchy of the data. The SQL fact tables are used in the MeasureGroups of the SSAS database and the aggregation logic (SUM,MIN,...) is used to determine how that field is used. You can also define the relationship between the fact data and the dim data in the Cube.
The first step of a SSAS solution would be define the SQL data warehouse. The SQL data warehouse would usually be a star schema dimensional data model populated from your source system data. Once you create the data warehouse with Fact tables and Dimension tables, you can start creating your SSAS database.
1. Create an SSAS solution. Right click on the projcet (databse) and you should have a popup with Deployment configuration. Make sure you are deploying to the SSAS server you have access to and define the SSAS database name you want it deployed as.
2. Create a data source connecting to the data warehouse. Make sure you configure the impersonation settings correctly for processing after deployment.
3. Create a data source view from the data source you just created and import the fact and dimension tables.
4. Create a dimension for each of the dimension tables.
5. Create a Cube with measure groups using the fact tables.
6. Go to the Dimensions tab in the Cube and make sure dimensions are mapped to the fact. (If the relationship between fact and dimension are defined in the dsv, the connection should have been automatically created.) If a dimension is missing, right click and add the dimension. Make sure fact and dimension relationships are defined.
7. Deploy and process.
What happens at this point is:
The SSAS database definition will attempt create/alter objects on the SSAS server and database you defined in step 1.
Once the objects are deployed, it will attempt to process the data. Processing of the data entails selecting the fact and dimension data from the data warehouse tables and aggregating and storing it in the SSAS database.
I hope this all makes sense and helps. It's a Saturday and my mind is not at 100%.