Write scalable SQL with Oracle Autonomous Database
summary
Shows three concrete antipatterns and fixes: concatenating user input causes SQL injection and heavy parsing — switching to bind variables turns 5,000 lookups from ~7s to <0.1s (≈118× faster); putting SQL inside loops wastes network and CPU — moving to PL/SQL FORALL with RETURNING batches 100,000 inserts from ~2.5s down to ~0.5s (≈8×) and cuts an 8.5s ship-loop to 3.5s; querying one table repeatedly becomes N× work — replacing scalar subqueries with analytic OVER ROWS drops a 13.5s run to 0.1s (≈135×) by avoiding 127M-row re-scans.
AI-written summary · about
chapters (6)
- 0:00Database tuning fundamentals
- 4:12Avoiding string concatenation
- 9:04Efficient SQL execution
- 13:38Eliminating loops in SQL
- 18:49Advanced bulk processing
- 22:34Optimizing analytical queries
related talks
52mExploring PL/SQL New Features and Best Practices for Better Performance
50mSupercharge your Code to Get Optimal Database Performance
47mHow to query data using natural language - intro to AI features in Oracle 23ai
42mEfficient Data Handling Techniques in Python for Oracle Database
36mHow to make your daily routine, code and database more efficient
21mDeep dive: Data loading
33mData handling techniques in Python for Database
31mFuturistic SQL: Making the impossible possible