Microsoft SQL Server 2017 Standard

Microsoft SQL Server 2017 Standard is the landmark release that made SQL Server a truly cross-platform database engine by introducing native Linux support for the first time in the product's history. Organizations running Ubuntu, Red Hat Enterprise Linux, or SUSE Linux can now deploy the full SQL Server engine with feature parity to Windows, opening Microsoft's database platform to the vast Linux server ecosystem.

Adaptive Query Processing represents a fundamental shift in how SQL Server optimizes queries at runtime. Instead of relying solely on compile-time estimates, the query processor now adjusts execution strategies during query execution based on actual row counts. Adaptive joins switch between nested loop and hash joins mid-execution, interleaved execution provides accurate cardinality estimates for multi-statement table-valued functions, and memory grant feedback automatically corrects oversized or undersized memory allocations on subsequent executions.

Graph database support adds native graph processing capabilities directly within SQL Server, enabling you to model and query complex many-to-many relationships using NODE and EDGE tables with MATCH syntax. Social networks, recommendation engines, fraud detection systems, and network topology analysis benefit from graph queries that would require complex recursive CTEs or multiple self-joins in traditional relational models.

SQL Server 2017 introduced in-database machine learning with R and Python integration. The Machine Learning Services feature executes R and Python scripts directly inside the database engine, bringing predictive models to the data instead of extracting data to external tools. This eliminates data movement latency and security risks associated with copying production data to analytical workstations.

Automatic database tuning continuously monitors query performance and automatically creates, validates, and applies covering indexes when it detects repeated performance regression patterns. The system can also force the last known good execution plan when a plan regression is detected, fixing performance issues without DBA intervention.

Resumable online index rebuild lets you pause and resume index maintenance operations across maintenance windows, eliminating the need to restart multi-hour index rebuilds from scratch when a maintenance window expires. This is invaluable for large tables where index maintenance cannot complete within a single overnight window.

SQL Server 2017 Standard supports up to 128 GB RAM and 24 cores, with extended support from Microsoft through October 2027.