📄️ Overview
Composite backends in Stalwart Mail Server provide a flexible way to combine multiple storage backends, allowing administrators to distribute load effectively or increase redundancy for critical data. By leveraging composite backends, Stalwart Mail Server can optimize performance and ensure resilience in environments with high availability and scalability requirements.
📄️ Sharded In-Memory Store
Sharding is a technique used to distribute data across multiple storage backends, improving performance and scalability by evenly spreading the load. In the context of Stalwart Mail Server's sharded in-memory store, sharding works by hashing the key of the data to be stored or retrieved and using a modulus operation to determine which in-memory store should handle the request. This ensures that each backend is responsible for a specific subset of the keys, enabling efficient use of resources and preventing bottlenecks.
📄️ Sharded Blob Store
Sharding is a method used to distribute data across multiple storage backends, enhancing scalability and performance by balancing the load among them. For Stalwart Mail Server's sharded blob store, sharding operates by hashing the key associated with a blob and using a modulus operation to determine which storage backend will store or retrieve the blob. This ensures that each backend manages a specific portion of the data, optimizing resource utilization and access speed.
📄️ SQL Read Replicas
SQL read replicas are a database architecture pattern used to improve performance and scalability by separating read and write operations. A primary database handles all write operations, ensuring that data is consistently updated and stored. Read replicas, on the other hand, are synchronized copies of the primary database and are used exclusively for read operations. This separation helps distribute the load, reducing strain on the primary database and enabling efficient handling of high-volume read queries.