We’d love to hear more from you at http://www.facebook.com/office

Cheers,
Andy
MSFT Office Outreach Team

If the manager is refreshing small datasets, then it’ll be fast and no one will really notice. If however the datasets are huge, then it will only take one time of the manager waiting around a while for his refresh to pull all that data to cure him of the “just to be safe” situation.

Further, if the manager does persist then he’ll not only have phone calls from irate DBAs, but he’ll have to deal with complaints from his own staff (or potentially other managers) for killing performance of production systems.

If / when the PowerPivot sheet is uploaded to Excel 2010, a schedule can be setup there to do the refreshes. Now this may actually mean a SharePoint Admin and a DBA might actually have to talk to each other, but stranger things have happened.

If all that doesn’t fix it, then I advise that a data warehouse, or even just a simple copy of the production data be done each night, and then only allow people to pull from the copy and not the live database. In the long run this is what I suspect will happen in many organizations.

Also recall that SQL Server Analysis Services can be used as a data source, so a very likely scenario would be a BI project is created that sets up an SSAS database, but instead of creating pre-built pivot tables users are allowed to create their own pivot tables / charts in PowerPivot using SSAS as the source.

And there is a silver lining, in that when a DBA sees a lot of hits from the same PowerPivot spreadsheet, IT will know it’s a popular sheet and perhaps some IT dollars should be focused there to create a full blown BI solution, as opposed to the “pointy haired boss” who is yelling loudly to the IT manager that his sheet needs to be converted although he’s the only one who uses it.