pganalyze in action: The Latest Features for Tuning Postgres - Webinar Cover Image

Webinar: pganalyze in action: The Latest Features for Tuning Postgres

The release of PostgreSQL 18 introduced significant changes that directly influence performance at scale: from the introduction of asynchronous I/O, which changes how Postgres interacts with the disk both in the cloud and on-premise, to new planner optimizations that can make queries faster. We examine these changes in detail, using demonstrations and examples to show how they can be applied to achieve faster queries, more predictable scaling, and improved observability in PostgreSQL environments.

Lukas Fittl
Founder & CEO, pganalyze

In detail, we walk through

Asynchronous I/O: A fundamental change in how Postgres handles I/O, offering the potential for significant performance gains, particularly in cloud environments where latency is often the bottleneck.
Indexing Improvements: How the new B-tree Skip Scan builds on Postgres 17’s improvements to more efficiently find a set of values, how it compares to loose index scans, and how indexing strategies should evolve.
UUIDv7 for Primary Keys: Understand why UUIDv7 improves performance and ordering, how to migrate from UUIDv4, and when bigint still makes sense.
Planner Changes: How Postgres 18 interacts with queries, for example, removing unnecessary self-joins, transforming OR-clauses into array scans, optimizing DISTINCT, and enabling more efficient join strategies.
VACUUM Enhancements: The new autovacuum_vacuum_max_threshold setting and how it helps optimize autovacuum schedules for large tables with many rows, and other ways that VACUUM behavior is changing.
Monitoring Upgrades: From richer EXPLAIN output to extended pg_stat_* views, learn how Postgres 18 makes it easier to understand query behavior, WAL activity, NUMA interactions with shared buffers, and more.
pg_stat_plans: The addition of the PlannedStmt.PlanID column in Postgres 18 enabled the ability to track plan-level metrics and query execution trends with our new open source extension, pg_stat_plans.
Other Performance Gains: Partition planning improvements, increased fast-path lock limits to reduce locking overhead for large-scale workloads.

Watch the Webinar Recording

Hundreds Of Companies Monitor Their Production PostgreSQL Databases With pganalyze