Azure Database Updates: Hot patching; SQL Database; Auto-failover groups; Data Factory; Cosmos DB composite indexes

After weeks without many database updates, Microsoft ramped up its database announcements as it reached the middle of September.

Principal software engineer Hans Olav Norheim went over hot patching for Azure SQL Database. The program promises a continually up to date SQL Engine, but this means that periodic updates must take place. Failover of a primary replica during a one-second stop in the SQL Server process stops transactions and in-flight queries. Although Microsoft keeps multiple replicas of every database, Norheim recommends that users do hot patching, by modifying in-memory code, modifying C++ code without restarting. Hot fixes can resolve up to 80 percent of SQL issues, which cuts down on workload disruption and results in faster fixes.

In recent months, Microsoft has continually emphasized the need for continuity and it focused on that theme again with generally available auto-failover groups to help create geographically redundant copies of SQL Database data. The system can also recover multiple related databases to a secondary region in the event of a service interruption in the primary region.

Azure's developers have enabled Data Factory to copy data into Azure Database for MySQL, using the Copy Activity feature for one-time data movements or scheduled pipeline activity. This offers more than 80 supported data sources. In fact, Copy Activity is supported for all forms of Azure SQL Database including Managed Instance, Data Warehouse and SQL Server. Users may need to adjust sink table schemas after data is ingested.