Postgres 18 just dropped with a host of goodies—think native UUIDv7 support, an “Over Explain” module for deeper query insights, composite index skip scans, and (most buzzworthy) a brand-new asynchronous I/O engine. In this episode, Hussein walks you through sync vs. async calls, recaps how Postgres 17 handled I/O, then dives into 18’s dual async modes—worker threads versus io_uring—and even shows how the fallback sync path still fits in.
Along the way you’ll catch the real-world challenges the team solved, get a peek at backend writer support and direct I/O plans, and see why async in Postgres isn’t “done” but already a huge step forward for performance buffs everywhere.
Watch on YouTube
Top comments (0)