SQL Server 2017 New BI Features SQL Saturday Charlotte October 14, 2017 David Mellinger
About me David Mellinger Data Platform & Analytics Practice Lead at DMI\Allegient Twitter: @HoosierDataDev LinkedIn: linkedin.com/in/dmellinger MCSE: BI\Data Management & Analytics Live in Columbus, Indiana Active member of Terrace Lake Church Sports fan HS, College, and Pros various sports
Agenda Overview of New BI Features in SQL Server 2017 SQL Server Reporting Services (SSRS) - Demo Power BI Report Server SQL Server Analysis Services (SSAS) - Demo Azure Analysis Services BISM Normalizer SQL Server Machine Learning Services (formerly R Services) - Demo SQL Server Integration Services (SSIS) Talk about more to SQL Server 2017 than just Linux 2017 has a decent amount of BI\Analytics features, particularly considering its 1 year out of a release This will likely become the norm, b\c cloud first dev allows quick turnaround and user testing in the cloud resulting in shorter cycles
SSRS Web Portal Report comments Comments can be added to all report types (Paginated Reports, Mobile Reports, and Power BI Reports) Can add picture or PDF attachments to comments Basic Reply and Security options available Adds some light collaboration to the Web Portal The only feature enhancement to SQL Server 2017 per se are comments Only other SQL Server 2017 specific feature is the lightweight stand alone installer and separate install path
SSDT and Report Builder Native DAX Support Native DAX queries in SSDT and Report Builder Graphic Designer works with SQL Server 2016+ Tabular Models Can switch to text and write DAX for 2012, 2014 Tabular models
Power BI Report Server Superset of SSRS, with the ability to surface Power BI Reports Secure and govern Power BI reports in SSRS Web Portal Power BI Desktop edition Currently limited to Live Connection to (Tabular or MD) Analysis Services models Surfacing Imported Models and Excel worksheets is in preview License comes with Power BI Premium or SQL Server Enterprise Edition with Software Assurance Secure and Govern More limitations to the visuals available ?rs:embed=true
SQL Server 2017: SSRS and power bi report server Demo
SSAS New Features Modern Get Data Experience Get Data M\Power Query Data Sources Power BI like data source experience Can still utilize “legacy” data sources Currently quite buggy in SSDT Many more data sources Lot of potential; easy to load and relate non-DW data, lookup sKey via natural key and tie ad-hoc data to EDW
SSAS New Features Object Level Security Object Level Security (Table and Column metadata security) Previous security was data security These features prevent users via roles from being able to see if a table or column even exists in the model This also closes loophole of seeing ‘hidden’ data via Show Details
SSAS New Features Custom Detail Rows for Measures Ragged Hierarchy Support (Hide Members) Encoding Hints Value: Better for aggregate, dense number fields Hash: Better for lookup\key, sparse number fields DAX Editor in SSDT SSMS DAX Query Editor
SSAS Related Features BISM Normalizer Free\Open Source SSAS Tabular Schema Compare Tool Should be core tool used for SSAS Tabular development and management Easy to see developers changes from deployed models or between environments Allows easy incremental deployments Allows for much easier concurrent development Can deploy model, but maintain Partitions, so development can be loaded and be manageable with much less data (different partitions) Now in SSAS Git Repository If you are going to take away anything from this presentation, this should be it If you have never heard of BISM normalizer or aren’t that familiar, I encourage you to dive into the tool, I promise it will help you out immensely
SSAS Related Features BISM Normalizer If you’ve ever done a TFS right-lick compare on Tabular Model, you know how ugly, long that is and how it is hard, sometimes impossible to compare
Azure Analysis Services Web Designer Can create models from Azure SQL DB, Azure SQL Data Warehouse, or a Power BI Desktop File Power BI Desktop importing via SSDT to SSAS Tabular project is on the roadmap
Azure Analysis Services Web Designer Intended for quick creations of basic models or creating simple changes Can check basic querying as well SSDT is still primary authoring tool Its intended for quick creation of models or applying simple changes quickly, of course you need to be careful with changes to production models
SQL Server 2017: SSAS and BISM Normalizer Demo
Machine Learning Services Python Support Formerly SQL Server R Services Renamed as it now supports Python and possibly supporting more platforms in the future In addition to In-Database options for running R\Python on an instance of SQL Server, there are also Machine Learning Server options RevoScalePY and MicrosoftML packages both support Python Python has same Data Science and Support as R has in SQL 2016
Machine Learning Services Native Scoring New PREDICT T-SQL function Takes a pretrained model and data and scores it Uses optimized intra-database C++ libraries, does not call separate R\Python runtime Fast performance No external script security concerns Train a model in Data Science environment, save serialized model to SQL Server table, then can be used to score via PREDICT No need for any Machine Learning Services to be installed on the SQL Server instance Currently only supported for a few RevoScaleR algorithms, likely RevoScalePY and MicrosoftML algorithms in the future
SQL Server 2017: Machine learning services Demo
SSIS New Features Scale Out SSIS Scale Out Allow packages from a SSIS Project Deployment to run on different servers concurrently Consists of one Master and one or more Workers which pull jobs from the Master Master server can also be a worker Can be ran via SSMS Dialog, via SSIS Stored Procedures, or via SQL Agent Will NOT run child packages called from Execute Package Tasks on separate Workers, so may require redesign to take advantage If use ELT a lot or if database is bottleneck, not too helpful Most helpful if SSIS server’s Memory or CPU is bottleneck
SSIS New Features SSIS Azure Feature Pack Updates Azure Data Lake Store Azure SQL Data Warehouse HDInsight Updates Linux Support (sort of) Severely limited Ubuntu and Red Hat only File System Deployment Only Call via dtexec No SQL Agent Support Dynamics Online ODBC connections support Azure Data Factory can now run SSIS Packages (any version)
Microsoft BI\Data Analytics Feature Comparison SQL Server 2017 Windows SQL Server 2017 Linux Azure\Cloud Data Warehouse SQL Server RDBMS Azure SQL DB Azure SQL Data Warehouse Data Movement SSIS SSIS (weak) Azure Data Factory Semantic Model SSAS (Tabular and MD) N\A Azure Analysis Services (Tabular) Reporting SSRS (Paginated, Mobile, Power BI Reports) Power BI (Service and Reports) Advanced Analytics SQL Server Machine Learning Services Azure Machine Learning Apache Spark for Azure HDInsight Master Data Management Master Data Services Big Data Support Interaction through Polybase Azure HDInsight Azure Data Lake
Resources Links What’s new in SQL Server 2017 BISM Normalizer Blog and Whitepaper Azure Analysis Services Web Designer Power BI Report Server Preview (Imported Models and Excel Surfacing) SQL Server 2017 download and play around Blogs, Google and explore!
Questions?