Reducto faced a challenge in managing its rapidly scaling customer usage due to a bottleneck in its Postgres instance on Amazon RDS, necessitating a zero-downtime database migration to avoid disrupting its 24/7 API services. After evaluating several options, including PlanetScale for Postgres, which was unsuitable due to its requirement for logical replication and mandatory backups conflicting with Reducto's Zero Data Retention policy, the company opted for PgDog. PgDog allowed Reducto to mirror traffic from the old to the new database instance without needing application changes or logical replication. The migration process involved testing PgDog's ability to handle production-scale client connections, monitoring mirroring metrics, and ensuring data consistency before swapping databases, all of which were meticulously planned and executed with the assistance of PgDog's team. The successful migration, which included a version upgrade from Postgres 16 to 17, was accomplished without downtime or data inconsistencies, showcasing PgDog's effectiveness and marking a significant operational improvement for Reducto.