KubeForge
A local Kubernetes project demonstrating microservices orchestration, Horizontal Pod Autoscaling (HPA), load generation, and comprehensive observability.
The Narrative
Architected a Minikube Kubernetes cluster to deploy custom Go microservices handling CPU-intensive workloads to simulate production constraints.
Implemented Horizontal Pod Autoscaling (HPA v2) to dynamically scale deployments from 1 to 10 pods based on real-time CPU utilization metrics.
Set up a comprehensive observability stack using Prometheus, Grafana, and kube-prometheus-stack for tracking pod metrics, network bandwidth, and IOPS.
Automated system load testing with k6 scripts simulating constant-arrival-rate traffic and managed cluster configuration using Helm and Kustomize.
Tags
Links
System Highlights
Orchestration
Managed Deployments, Services, and path-based Ingress routing within a local Kubernetes cluster.
Autoscaling
Leveraged metrics-server and HPA to automatically scale Go service replicas under synthetic load.
Observability
Deployed Prometheus and Grafana for real-time monitoring of CPU usage, network I/O, and pod lifecycle.
Go Workers
Engineered highly concurrent zero-dependency Go microservices to intentionally saturate CPU limits.
