This repository contains tools and configuration files for the testing, benchmarking and automation needs of the databend projects.
- Fast CI speed, By design one performance testing should not exceed two hours(including docker build time and performance testing running time)
- Expandable: supports to deploy performance tests on scale, and also supports to deploy on a single machine for affordable CI
- Cloud Native supports: Should be able to deploy whole platform on different cloud providers like GKE, EKS
- High Availability: supports server instances self-healing and do not have single point failure
- Observability: whole process should be observable, should collect logs for performance running instances and collect compare report results
- Cloud provider interface implementation
- dashboard for unified monitoring
- prometheus and grafana integration
Steps:
- Chatbot receive PR comments for performance testing
- Chatbot validate on commenter’s permission
- Trigger performance docker image build for current and reference instance
- Create new cluster for testing
- Deploy instances and performance tool on it
- Run performance testings and collect performance results
- Clean up and delete all created resources(cluster and docker image layer)