Most of the world's data exists in a combination of on premise, cloud and non-relational forms. What you normally do is use a tool like SSIS to export from Teradata, Oracle, SQL Server, Flat Files, Hadoop, or some other source and then load it into a SQL Server database. You may use Cubes, you apply indexes, then there are visualizations that are constructed. That was your data warehouse. Now you are being asked to provide more data from different sources, and people want to run something called R against the data. Enter SQL Server 2016. You have the ability to connect to multiple data sources using Polybase, to have SQL Server treat those tables not just as linked servers but first class citizens with Polybase. You have In-database R, the ability to install R packages, and in vNext you have on premise Power BI. In this session we will review what the new Modern Data Warehouse looks like, how we ingest data, review data, connect R Studio to SQL Server, move R scripts to In-Database R, and visualize the data with Power BI.
You will learn:
- How to create an external data source and use Polybase to query Hadoop data
- How to install R packages to SQL Server, Connect R Studio to SQL Server, and create an In-Database R stored procedure
- How to visualize the database output with Power BI in Azure and on premise