AWS recently announced readable standby instances in Amazon RDS-Multi-AZ deployments. The new database cluster option provides high availability and durability for workloads that require more read capacity and consistent transaction latency.
Introduced more than ten years after Multi-AZ deployments for RDS, the new preview option uses native replication of the database engine and is based on local NVMe SSD storage and R6gd / M6gd Graviton2 instances. The standby instances act as automatic failover targets and can serve read traffic to increase throughput without adding read replicas.
The new approach overcomes one of the limitations of traditional RDS Multi-AZ deployment where the synchronously replicated standby cannot accept connections. Vijay Karumajji, Database Solutions Architect at AWS, explains how readable standby instances work:
A Multi-AZ-DB cluster consists of one primary and two readable standby read replicas in three different …