MariaDB Server must adapt seamlessly to hardware resource changes (CPU, RAM) in dynamic environments like Kubernetes/Cloud, where allocations can shift online without server restarts. This story covers the development of automatic scaling mechanisms to optimize thread pool and buffer pool usage based on real-time hardware availability, reducing manual tuning and enhancing cloud-native performance.

Current State

1. Thread Pool Dynamism: The thread pool adjusts active threads based on workload within thread_pool_min_threads and thread_pool_max_threads.
2. Manual CPU Scaling: thread_pool_size can be changed via SET GLOBAL, but there’s no automatic adjustment for CPU changes.
3. Buffer Pool Static Configuration: innodb_buffer_pool_size defaults to 128MB and requires manual tuning, despite recommendations of 75-80% of RAM.
4. Kubernetes & Cloud Constraints: Containerized environments dynamically scale CPU and RAM, but MariaDB relies on manual intervention or scripts to optimize resource usage.

*Proposed Feature Set: Automatic CPU-Based, RAM-Based Auto-Scaling and Hill-Climbing (Enterprise).
Introduce configurable auto-scaling mechanisms to dynamically adjust the thread pool and buffer pool based on resource changes.