This article discusses a system built for Wayfair that uses PostgreSQL as a Dead Letter Queue (DLQ) to manage failed event processing. Instead of using Kafka for failed events, the system stores them in a PostgreSQL table, allowing for better visibility and easier reprocessing. It also outlines a retry mechanism with exponential backoff to prevent flooding the DLQ with transient failures.

This article details Datadog's approach to creating a managed data replication platform that improves data movement across services. It covers technical challenges faced with a shared Postgres database, the transition to a dedicated search platform, and the use of asynchronous replication to enhance scalability and reliability.

Fresha successfully executed a zero-downtime upgrade from PostgreSQL 12 to 17 across over 200 databases by developing a tailored upgrade framework that addressed the complexities of maintaining data consistency and availability during the process. The approach involved leveraging logical replication, managing Debezium connectors, and implementing a two-phase switchover to ensure a seamless transition without disrupting production services.