Modern cloud applications place new demands on PostgreSQL: fast and predictable failover, read scalability without operational complexity, and continuous availability under dynamic workloads. Meeting these requirements requires rethinking not just deployment, but core architectural and operational assumptions.
In this session, we take an engineering‑centric look at HorizonDB, a new managed PostgreSQL service on Azure, and the design decisions behind it. We will explore how HorizonDB approaches high availability, immediate failover, and read replication, and how these capabilities are delivered consistently without exposing application teams to operational burden.
This session is aimed at engineers and architects building or operating large‑scale PostgreSQL systems who want to understand the trade‑offs, constraints, and practical lessons involved in designing a next‑generation PostgreSQL platform for the cloud.