AutoMQ has been redesigned based on a shared storage architecture to implement Kafka. The cost advantages over Apache Kafka come from following aspects:
Storage Cost Optimization:
The storage layer uses object storage as the storage service, which costs one-tenth of regular cloud disks. It also eliminates the need for triple replication, allowing for on-demand usage and pay-as-you-go pricing.
Compute Cost Optimization:
AutoMQ separates data persistence to cloud services, removing the need for upper-layer triple replication. This reduces replication traffic and pressure by two-thirds, enabling the same machine resources to support a larger scale of data read/write operations.
Auto-Scaling:
AutoMQ's stateless architecture supports auto-scaling. With AutoMQ, there is no idle costs; it is fully on-demand and pay-as-you-go.