pkanderi-abio/vmware-dashboard

Purpose and Entry Points

VMware Dashboard is a full-stack operations dashboard for monitoring multiple vCenter environments and enriching VM inventory with historical CMDB and optional PuppetDB data. The backend FastAPI application starts at main.py; the React/Vite SPA starts at src/main.tsx and composes routes in src/App.tsx.

Backend API and Application Boundary

main.py is the API boundary for browser clients: it exposes inventory, configuration, health, search, action, CMDB, and integration endpoints, applies CORS, and coordinates backend services. Browser API targeting is centralized in src/config/api.ts and consumed through src/lib/api.ts; the frontend supports a runtime backend URL override rather than requiring rebuilds.

vCenter Connection and Collection Pipeline

vCenter endpoints and server-local credentials are managed through the backend, with credentials persisted outside the repository under the user cache directory. Parallel pyVmomi PropertyCollector-based inventory gathering is implemented in vcenter_fast.py; related VMware domains are separated into lifecycle, tags, certificate, and health modules:

• vcenter_fast.py

• vcenter_health.py

• vcenter_lifecycle.py

• vcenter_tags.py

• vcenter_certs.py

The collector is the core external-system boundary to vCenter over HTTPS/443 and supplies hosts, VMs, datastores, networks, and snapshots for API responses.

Inventory, Cache, and Operational Actions

The backend converts collected vCenter state into dashboard resources, supports cached refresh behavior, and retains historical CMDB state for decommission tracking. cmdb_history.py owns inventory history; global_search.py provides cross-resource search; action_guard.py gates operational actions before they reach VMware-facing logic. Snapshot and VM action affordances are surfaced from:

• src/components/vm-actions.tsx

• src/components/snapshot-actions.tsx

• src/pages/snapshots.tsx

CMDB and Puppet Enrichment

CMDB records combine vCenter VM facts with optional PuppetDB data, allowing detail views and exports to show operational classification, facts, and history. puppet_client.py is the PuppetDB REST boundary, while cmdb_history.py persists local historical state. Primary UI consumers are:

• src/pages/cmdb.tsx

• src/pages/vm-detail.tsx

• puppet_client.py

Frontend Presentation and Navigation

The SPA uses React 18, TypeScript, React Router, Tailwind, and shadcn-style components. The app shell/sidebar and global search live in src/pages/_layout.tsx; resource pages render independent inventory perspectives, notably src/pages/index.tsx, src/pages/vms.tsx, src/pages/hosts.tsx, and src/pages/observability.tsx. Theme state persists client-side via src/lib/theme.tsx.

Deployment and Runtime Infrastructure

Backend runtime is Python 3.9+ with FastAPI/Uvicorn and pyVmomi; frontend is built by Vite and served as static assets or behind a reverse proxy. Container builds are split by tier in docker/api.Dockerfile and docker/frontend.Dockerfile. Kubernetes manifests define API/frontend deployments, ingress, persistent storage, HPA, network policy, secrets, and an in-cluster runner:

• k8s/kustomization.yaml

• k8s/api-deployment.yaml

• k8s/frontend-deployment.yaml

• k8s/runner/deployment.yaml

GitHub Actions validates frontend builds and supports self-hosted-runner deployment through .github/workflows/ci.yml and .github/workflows/deploy.yml.