Amazon Aurora Serverless is a fully managed, auto-scaling configuration of the Amazon Aurora database that automatically adjusts computing capacity based on the application's actual needs. Aurora is much like other database engines but with significant differences in performance, scalability, and availability due to its integration with the Amazon Web Services (AWS) infrastructure.
Aurora Serverless v2 supports the Aurora I/O-Optimized cluster configuration for I/O-intensive applications, as an alternative to Aurora Standard.
The serverless aspect of Aurora allows users to create a database without the hassle of managing complex and time-consuming administrative tasks such as hardware provisioning, software patching, setup, configuration, or backups. Instead, Aurora Serverless handles these operations behind the scenes, offering a seamless experience for the developer or database administrator.
One of the core benefits of Aurora Serverless is its ability to automatically start-up, shut down, and scale the database capacity with the flow of the application traffic. This is a significant advantage for workloads that have unpredictable or variable database usage patterns. For example, a database that serves backend services for a retail application might need to scale up during busy shopping periods and scale down overnight or during quieter periods to save costs. By automatically adjusting compute capacity in real time, without any disruption to client connections or application performance, Aurora Serverless ensures users pay only for the database resources they actually use.
Aurora Serverless v2 scales within a capacity range you specify, supporting up to 256 Aurora Capacity Units (ACUs) for Aurora MySQL and 384 ACUs for Aurora PostgreSQL. The available range depends on the engine and platform versions, with supported versions allowing a minimum of 0 ACUs. Each ACU combines approximately 2 GiB of memory with corresponding CPU and networking.
Aurora clusters can combine provisioned and serverless instances in a mixed configuration, including by adding serverless instances to an existing provisioned cluster.
The capacity range is configured at the cluster level and applies to every serverless writer and reader, but each instance has its own capacity within that range.
The default max_connections setting is calculated using the memory associated with maximum configured ACUs rather than current capacity. Changing the maximum capacity requires rebooting the serverless instances to update this static parameter.
Capacity-range changes take effect immediately, even if you select the next scheduled maintenance window.
To maintain high availability and durability, Aurora Serverless spreads data across multiple AWS Availability Zones - independent infrastructure in separate geographic locations. Data is replicated six times across three availability zones, ensuring that the loss of a single instance or even an entire data center, does not lead to data loss or downtime.
Despite the 'serverless' name, there are indeed servers in the background; the serverless aspect refers to the abstraction for the user. Developers can simply interact with the database endpoint, and the service manages the compute resources. This means that tasks such as capacity planning, which would typically require precise prediction and careful resource allocation considerations, are handled automatically and optimized to match real-time demand.
Aurora Serverless integrates with AWS services like AWS Lambda, permitting users to trigger functions in response to database events, thus creating a reactive, event-driven architectural pattern. It also supports industry-standard SQL, making it compatible with existing applications and tools. Security in Aurora Serverless is not an afterthought. The service includes network isolation using Amazon VPC, encryption at rest using AWS Key Management Service (KMS), and encryption in transit with SSL. Access management is handled through AWS Identity and Access Management (IAM) roles and policies, while security patches are applied automatically.
In essence, Amazon Aurora Serverless brings forth an efficient, cost-effective, highly available, and secure database solution ideal for many types of applications, especially those with variable workloads and the need for simplifying operations. It provides the performance and availability of high-end commercial databases with the simplicity and cost-effectiveness of open-source databases, all while leveraging the advantages of the AWS cloud infrastructure.
