Right-Sizing and Auto-Scaling of MySQL Containers in Kubernetes
summary
An automated system measures MySQL container CPU/memory usage, runs statistical forecasting, and runs an optimizer that adjusts request&limit via Kubernetes and Vitess APIs and triggers replica-based migrations to avoid restarts and preserve QoS. Scheduling uses a multi-resource similarity matching score that weighs CPU, memory and migration cost to pick target hosts. Field results show successful placement of tens of thousands of containers and about a 20% resource reduction while keeping performance within targets.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:57JD Elastic Database
- 2:49Problems and Challenges
- 4:06Workload Characterization
- 5:30Right Sizing Estimate the worldoad demand
- 7:20Right Sizing of CPU Resources
- 16:17Auto Scaling and Rescheduling
- 18:34Auto-Scaling: Overview and Cost Models
- 20:08Host Selection: Multi-Resource Balance
- 24:02Experimental Evaluation Setup
- 25:50Host Selection: Resource Availability
- 28:35Host Selection: Correlation-awareness
- 29:15Key Results
- 30:55Conclusions
related talks
36mRight-Sizing and Auto-Scaling of MySQL Containers in Kubernetes
35mRunning MySQL on Kubernetes [I]
29mRight-Size Your Kubernetes Deployments
31mFailing to Autoscale?
14mStop Hand Tuning MySQL Teach Kubernetes Operator
41mDemocratizing MySQL: From Cloud Managed to Kubernetes Managed
23mRunning MySQL on K8s. Deployment, Scaling, Backups and advanced configuration
31mVitess: Running MySQL with confidence on Kubernetes