PSAP Control Center

A cluster management and reservation platform for the Performance and Scale for AI Platforms (PSAP) team. Provides observability into OpenShift cluster health, a reservation system with calendar views, kubeconfig management, and Hearth GPU discovery integration.

Live instance: control-center.apps.psap-automation.ibm.rhperfscale.org

Features

• Cluster Registry — Add clusters via kubeconfig upload or kubeadmin credentials. Live health monitoring, node topology visualization, OCP details, operators, and workloads.
• Reservation System — Full-cluster or partial GPU reservations with type-aware conflict detection. Color-coded calendar, cancellation tracking, and historical preservation when clusters are removed.
• GPU Allocation & DRA — Live GPU status via DRA (resource.k8s.io) with automatic fallback to legacy node capacity counting. Per-GPU-type breakdowns, ConfigMap-driven vendor abstraction.
• Namespace Enforcement — GPU reservations automatically provision isolated Kubernetes namespaces with ResourceQuota and optional DRA ResourceClaimTemplate. Namespaces are cleaned up on completion, cancellation, or deletion.
• Calendar Views — Weekly preview with overlapping reservation display (index-based opacity for distinguishing coexisting reservations), plus full month/week/day calendar.
• Hearth Integration — Connect to a Hearth management cluster to discover GPU inventory via FournosCluster CRDs.
• View-Only Public Access — Anyone can browse clusters, reservations, and status. Modifications require sign-in.
• Structured Logging — Consistent log format across backend and frontend, configurable via LOG_LEVEL.

Tech Stack

Layer	Technology
Frontend	React 18, TypeScript, Vite, Tailwind CSS, TanStack Query
Backend	Python, FastAPI, SQLAlchemy (async), Kubernetes client
Database	SQLite (dev) / PostgreSQL (prod)
Deploy	OpenShift (binary builds), Docker Compose (local)

Quick Start

Docker Compose

git clone https://github.com/openshift-psap/psap-control-center.git
cd psap-control-center
cp .env.example .env    # Edit credentials as needed
docker-compose up -d

• UI: http://localhost:3000
• API docs: http://localhost:8000/docs

Local Development

# Backend
cd backend
python -m venv venv && source venv/bin/activate
python -m pip install -r requirements.txt
ADMIN_USERNAME=admin ADMIN_PASSWORD=admin \
  uvicorn app.main:app --reload

# Frontend (separate terminal)
cd frontend
npm install
npm run dev

• Frontend: http://localhost:3000 (Vite proxies /api to :8000)
• Backend: http://localhost:8000

OpenShift Deployment

See deploy/README.md for the full OCP deployment guide.

Authentication

Action	Auth Required
Viewing clusters, reservations, calendar, status	No
Creating, editing, deleting clusters or reservations	Yes (HTTP Basic Auth)

Credentials are configured via ADMIN_USERNAME and ADMIN_PASSWORD environment variables. On OCP these are stored in a Kubernetes Secret.

Sign in via the Sign In button in the top-right corner of the UI.

Configuration

Variable	Default	Description
ADMIN_USERNAME	admin	Admin username for write access
ADMIN_PASSWORD	admin	Admin password for write access
DATABASE_URL	sqlite+aiosqlite:///./psap...db	Async database connection string
SECRET_KEY	(change in production)	Application secret key
KUBECONFIG_STORAGE_PATH	./kubeconfigs	Directory for kubeconfig files
LOG_LEVEL	INFO	Backend log level (ERROR/WARN/INFO/DEBUG)
MLFLOW_BASE_URL	(optional)	MLFlow server URL
VITE_LOG_LEVEL	INFO	Frontend log level

Documentation

Document	Description
Architecture	System design, data model, API surface, deployment topology
User Guide	How to use the application
Contributing	Branch workflow, code standards, PR process
Troubleshooting	Common issues and solutions
OCP Deployment	Step-by-step OpenShift deployment guide

API

Interactive documentation is available at /docs (Swagger) and /redoc when the backend is running.

Key endpoints:

Endpoint	Method	Auth	Description
/api/v1/clusters	GET	No	List clusters
/api/v1/clusters	POST	Yes	Add a cluster
/api/v1/clusters/{id}/topology	GET	No	Node topology
/api/v1/clusters/{id}/gpu-status	GET	No	Live GPU allocation (DRA/legacy)
/api/v1/reservations	GET	No	List reservations
/api/v1/reservations	POST	Yes	Create reservation (full-cluster or GPU)
/api/v1/reservations/{id}	PUT	Yes	Update reservation
/api/v1/reservations/{id}/cancel	POST	Yes	Cancel reservation
/api/v1/reservations/calendar	GET	No	Calendar events
/api/v1/reservations/cluster/{id}/current	GET	No	Multi-occupant cluster occupancy
/api/v1/hearth/clusters	GET	No	Hearth GPU inventory
/api/v1/auth/check	GET	Yes	Verify credentials

See docs/ARCHITECTURE.md for the complete API reference.

License

Apache License 2.0

Related Projects

• TOPSAIL — Test Orchestrator for Performance and Scalability of AI pLatforms
• Performance Dashboard — RHAIIS benchmark analysis