Microsoft SQL Server 2017 Enterprise

Product information “Microsoft SQL Server 2017 Enterprise”

What’s new in SQL Server 2017

SQL Server 2017 is an important step in making SQL Server a platform where you have a variety of choices between different development languages and data types, between local execution or execution in the cloud, and between different operating systems. It provides improved performance of SQL Server for Linux, Linux-based Docker container and Windows. The following summarizes what’s new in specific functional areas and provides links with additional details.

SQL Server 2017 Database Engine

SQL Server 2017 includes many new database engine features, enhancements, and performance improvements. The database engine is the core service for storing, processing and backing up data. The database engine provides controlled access and transaction processing to meet the needs of the most demanding data processing applications in your business. Use the Database Engine to create relational databases for Online Transactional Processing (OLTP) or Online Analytical Processing (OLAP) data. This includes creating tables for data storage and database objects, such as indexes, views, and stored procedures for viewing, managing, and backing up data. Use SQL Server Management Studio to manage the database objects and SQL Server Profiler to record server events.

Machine Learning in SQL Server 2017

The SQL Server R services were renamed SQL Server Machine Learning Services, which now also supports Python, in addition to the R programming language. You can use machine-learning services (in-database) to run R or Python scripts in SQL Servers, or you can install the Microsoft Machine Learning Server (standalone) to deploy and use non-SQL R and Python models which require a server.

SQL Server developers now have access to the rich ML and AI libraries for Python, which is available in the open-source environment along with the latest innovations from Microsoft:

Revoscalepy: This Python equivalent to RevoScaleR includes parallel algorithms for linear and logistic regressions, decision trees, augmented structures, as well as a rich set of APIs for the transmission and movement of data, remote compute contexts and data sources.
Microsoft: This state-of-the-art package of Python-bound ML algorithms and transformations includes deep neural networks, fast decision structures, decision trees, and optimized algorithms for linear and logistic regressions. You’ll also get pre-defined models based on ResNet models that you can use for image extraction or viewpoint analysis.
Python in computer contexts of SQL Servers: Data analysts and developers can run Python code remotely from their development environment to try out data and development models without moving data.
Native Evaluation: The PREDICT function in Transact-SQL can be used in any instance of SQL Server 2017 to perform reviews, even if R is not installed. All you have to do is train the model with one of the supported RevoScaleR, algorithms and save it in a new, compact binary format.
Package management: T-SQL now supports the CREATE EXTERNAL LIBRARY statement to provide database administrators with better R package management capabilities. Use roles to control access to private or shared packages, save R packages to the database, and share them with users.

Advanced in-database analytics

Develop intelligent applications using SQL Server Machine Learning Services with R and Python. Switch from reactive to predictive and prescriptive analytics by running Advanced Analytics directly in the database. Use multithreading and massively parallel processing to get insights faster than with open-source R and python alone.

High availability and disaster recovery

Always On in SQL Server 2017 ensures high availability and disaster recovery on Linux and Windows. This provides you with less downtime, fast recovery, easier setup and load balancing on readable secondary replicas. In addition, it is possible to integrate an asynchronous replica into Azure virtual machines for hybrid high availability.