We’re pleased to announce support for Azure Analysis Services for Power BI solution templates. Effective today, we support Azure Analysis Services for the Campaign/Brand Management for Twitter, System Center Configuration Manager, Sales Management for Dynamics 365, and Sales Management for Salesforce solution templates.
Power BI solution templates simplify and accelerate building analytics solutions on popular applications, many that you probably use today. They offer a very quick guided experience to create compelling analytics and visualizations on an extensible, scalable, and secure architecture that offer immediate value and that you can customize as you see fit. This means that instead of spending weeks or months getting going, you can get started immediately and spend your time on extending and customizing the result to meet your organization’s needs. Learn more about Power BI solution templates.
So what is Azure Analysis Services and why should I care?
Well – first some background. Azure AS (let’s call it AAS) is another incarnation of the same engine that runs in Power BI Desktop, the Power BI service, and in SQL Server Analysis Services. Azure Analysis Services was announced several months ago in an article on the Azure blog. (Read it – good stuff.)
Now, why should you care about AAS? The Power BI Service is an amazing thing – why would you want to store your data in AAS rather than leaving it in Power BI? Especially when there is an added cost to do so.
The simplest and most obvious answer is data volume. The maximum size of a power BI desktop file that can be published to the Power BI Service is 1GB, after compression. For larger databases, you’ll need to bring in Azure Analysis Services.
This is a pretty simplistic way to look at things and AAS brings enterprise-ready capabilities that you might need long before you hit this data size limit. Size is a leading indicator, but here are some other things to consider.
Processing – How often do you want to refresh your reports? A report published to the Power BI service can be refreshed several times per day. If you want your model to be processed more frequently or you need more control on how it is processed, then AAS is one of your options. Reports bound to AAS are as fresh as the data inside it.
Partitioning – A table can be divided into logical parts each of which can be processed independently of the other. We don’t exploit this capability directly with solution templates (at least not yet) but you can. Solution templates are designed to be extended. So, for example, if you anticipate marrying your own data with what we provide, this can be important.
Client tools – As hard as this is to say, Power BI might not meet all your needs. Other tools like Excel offer some capabilities that Power BI does not. You might even have to use a *cough* competitor product. Azure Analysis Services supports not only Excel, but too many other client tools to mention here.
Size – Yes, it’s simplistic but it matters. If your model gets large you may need to turn to AAS to give you the Azure resources you need to not only host the data you have, but give you the performance you need.
So please try it out. As always, we look forward to your reaction and feedback. You can comment on your community site or simply email us at PBISolnTemplates@microsoft.com.