Presented by:

Sukhpreet Bedi

Amazon Web Services

Sukhpreet is a Senior PostgreSQL Specialist Solutions Architect at AWS, where she specializes in Amazon RDS and Aurora PostgreSQL engines. As a trusted technical advisor, she guides organizations in building resilient and secure database architectures on AWS. Her expertise extends to implementing Agentic AI solutions on PostgreSQL, helping customers develop intelligent applications that transform their business capabilities through modern database technologies.

Angel Duenas Neyra

Amazon Web Services

Angel Duenas is a Technical Account Manager at AWS, specializing in PostgreSQL. As part of the Technical Field Community, he guides customers in optimizing their PostgreSQL implementations on AWS. Angel helps clients build scalable, secure, and high-performance database solutions, enabling them to innovate and achieve their business objectives efficiently.

No video of the event yet, sorry!

Vacuum is the unsung hero of PostgreSQL, ensuring database health, reclaiming storage, and maintaining query performance. However, its internal mechanisms, recent improvements, and peculiar corner cases often remain a mystery to many users and developers. This presentation will explore the intricate workings of PostgreSQL's vacuum process, shedding light on:

  • Key architectural updates in vacuum internals over recent versions.
  • Optimizations that boost efficiency and reduce resource contention.
  • Handling challenging scenarios, including bloated tables, high-churn workloads, vacuum blockers and edge cases that defy conventional solutions.
  • Strategies for tuning and monitoring vacuum processes to suit diverse use cases.

Attendees will gain practical insights into leveraging vacuum enhancements while avoiding common pitfalls. Whether you're a developer, DBA, or performance enthusiast, this session promises to enrich your understanding of PostgreSQL's vacuum, its evolution, and its role in maintaining robust database systems.

Date:
2025 March 21 09:10 EDT
Duration:
50 min
Room:
Seminole E
Conference:
Postgres Conference 2025
Language:
Track:
Essentials
Difficulty:
Medium