The End of ANALYZE Hell: PostgreSQL 18's Statistics Preservation Story
Presented by:
Rajkumar Raghuwanshi
Rajkumar Raghuwanshi is a Database Consultant with AWS Professional Services based out of Pune, India. With good knowledge on relational databases adding hands-on in homogenous and heterogenous database migrations, helps customers in migrating to AWS cloud and their optimisations.
Nishad Mankar
Nishad Mankar is a Delivery Consultant with AWS Professional Services in India. He plays pivotal role in helping customers to migrate and modernize their databases on AWS cloud platform. With expertise in database technologies such as SQL Server, SAP ASE, and PostgreSQL, he builds cutting-edge solutions that helps customer success.
No video of the event yet, sorry!
PostgreSQL 18 introduces a game-changing feature that significantly improves the database upgrade experience by preserving optimizer statistics during major version upgrades. This talk will dive deep into the technical implementation and benefits of statistics retention, focusing on how it eliminates the need for lengthy post-upgrade ANALYZE operations that were previously required. We'll explore:
- The new architecture for statistics preservation using pg_dump/restore
- How pg_upgrade leverages this feature to maintain performance during version transitions
- Integration with vacuumdb's new —missing-stats-only option for handling edge cases
- Real-world impact on upgrade times and performance stability
- Implementation details of pg_stats_export and related infrastructure
- Limitations and considerations, particularly around extended statistics
This presentation will benefit DBAs, system administrators, and developers who manage PostgreSQL deployments and want to understand how to leverage this new feature for smoother, faster upgrades with minimal performance impact.
- Date:
- Duration:
- 25 min
- Room:
- Conference:
- PGConf India, 2026
- Language:
- Track:
- Database Administration
- Difficulty:
- Medium