This article explains how to use Oracle Database In-Memory on Amazon RDS for Oracle to improve real-time analytics performance without impacting OLTP workloads.

• Oracle Database In-Memory uses columnar format storage for faster analytical queries on large datasets
• Separate In-Memory Area (SGA) stores data in both row and columnar formats simultaneously
• Enables 78.62% faster OLAP queries while maintaining sub-second OLTP transaction times
• Requires Enterprise Edition BYOL licensing and minimum INMEMORY_SIZE parameter of 100 MB
• Use Performance Insights to identify suitable mixed workloads with IO-bound analytical queries
• Population can be on-demand (via full table scan) or priority-based (CRITICAL/HIGH/LOW)
• Automatic repopulation handles DML changes via transaction journals and threshold-based triggers
• Memory-optimized instance types (r5, r6) recommended for larger In-Memory workloads
• Real-world testing showed 86.49% response time improvement with concurrent OLTP and OLAP workloads

Oracle Database In-Memory provides significant performance gains for mixed workloads on RDS for Oracle with minimal application changes, requiring only parameter configuration and table attribute modifications.