Is it possible -- and how straightforward is it -- to migrate an existing AS2005 Cube (and/or at least it's dimensions) to an updated version of my source DB (SQL 2005)? I had little choice but to establish a "dev-and-test" version of the source DB during enhancements that coincided with production analysis and reporting. Now that I've approved the "dev-and-test" version, I want to make it our production master for OLAP.
As a note, I don't believe that any of the fields used in the cube have been modified. We've just added some new calculated fields (and sp's).
I have little idea of whether this will be simple, hard or impossible. I'm planning to try, but wanted expert feedback first.
There is a feature in BI Dev Studio specifically designed for the scenario you describe.
Open up your DSV and right click on the empty space anywhere. Select a "Refresh" option.
At this moment BI Dev Studio connects to the relational database and verifies if all the tables and all the coluns you build your Analysis Services objects upon are there. If it detects any changes in the relational database it will give you the list of the objects and will ask you if you would like that BI Dev studio will fix them for you automatically.
If the case is simple, just go ahead. If you see lots of changes, you save the list and try to takle every problem separately.
Hope that helps.
Edward Melomed.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
|||
Yes, the DSV refresh will help. However, just to clarify my situation...
I need to disconnect an existing cube away from it's source DB and connect it to a new database, (which is a modified, renamed copy of the original), not just refresh the DSV connection after some mods.
Two questions:
(1) In light of my situation,then prior to following your guidance above, would it work for me to (1st b/u everything, then) restore the new database into/overwriting the old database?
(2) Is there a better way involving DB file renaming, etc.? (Specifically, I'm looking for alternatives to re-building my cube and/or dimensions.)
I think I'm half way to resolving this problem. Thank you, Edward!
|||
Again, DSV refresh not only going to synchronize your DSV with relational database, it also checks and changes properties of attributes and measures. You should try it to see what it does for you and if it works.
Make a copy of your database and try your experiments on the copy.
You can also make backup of your database and then resore it with different name.
You can script your database Create script and modify the script to change the database name and ID.
You have quite a few options here. I suggest you just try and few and see what works.
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
|||
Thanks Edward. Your replies are illuminating and helpful.
In light of a new SSIS package connected to the new "dev-and-test" db, I'd like to inquire about simply changing the data source for an existing AS2005 cube, but will start a new thread on that subject.
