How ishaannene/blunder-buss works
Project type and purpose
• From the file tree we see a “web” folder containing HTML, CSS, and JS, a Go‐based “api” (go/api), a Go “worker” (go/worker), a containerized Stockfish engine (docker/stockfish), a Redis deployment in k8s, plus Terraform and k8s manifests.
• This is a full-stack, microservices-style chess move analysis platform:
• A static web UI served via Nginx (docker/web)
• A Go-based HTTP API (docker/api → go/api)
• A Redis-backed job queue
• A worker service that pulls jobs from Redis, invokes Stockfish (docker/stockfish), and writes results back
• Optional Go CLI client (go/client) for integration or testing
• All services are containerized, orchestrated via Kubernetes, and provisioned via Terraform.
File structure
• Top level:
• web/ → static frontend assets
• go/api, go/worker, go/client → three Go binaries (API server, asynchronous worker, CLI client)
• docker/… → Dockerfiles for each component plus an Nginx config
• k8s/… → Deployments, Services, HPA, KEDA ScaledObject for queue autoscaling, Redis PVC
• terraform/… → Infrastructure as code (e.g., EKS, VPC, IAM, kubeconfig template)
• scripts/… → build & push container images, local‐deploy scripts, API tests
• Patterns:
• Microservice decomposition (API vs. worker vs. engine)
• Queue‐based asynchronous processing
• Container per component
• Infrastructure‐as‐code with Terraform + K8s
Key insights from (empty) README
• Although the README is blank, the folder names and manifests convey the high-level architecture and dependencies (Go, Redis, Stockfish, Nginx, Kubernetes, Terraform, KEDA).