Skip to content

Security: openinferencefabric/oif

docs/SECURITY.md

Security

Reporting a vulnerability

Please do not open a public issue for a security problem. Report it privately through GitHub: Security > Report a vulnerability on https://github.com/openinferencefabric/oif/security. We aim to reply within 7 days and will credit you in the fix unless you ask us not to.

Problems in Run:ai, the KAI Scheduler or another upstream project belong to that project; if you are not sure, report it here and we will pass it on.

Requirements

Security requirements from the start, not before release.

Secrets

  • Secrets (Run:ai admin password, Grafana admin password) are read from .env at install time and stored only as Kubernetes Secrets. .env is untracked.
  • Scripts never echo secrets; logs are checked for token patterns in CI.
  • The repository is public: no real IP addresses, host names, user names or tokens in tracked files. CI runs gitleaks on every push.

Access

  • Kubernetes API and admin kubeconfig: readable by root and the installing user only (mode 600).
  • Run:ai: role-based access; the default admin password is changed at first login; projects grant only the users who need them.
  • Grafana: admin password from .env; anonymous access off.
  • Inference endpoint (v0.1): reachable only on the load-balancer address inside the lab network; no public exposure. Authentication at a gateway (API keys, rate limits) is planned for a later version and must exist before any exposure outside the lab.

Supply chain

  • Every image and chart pinned by version in versions.env; image digests recorded with benchmark results.
  • Images come from the publishers' registries (nvcr.io, docker.io/rocm, quay.io, registry.k8s.io, ghcr.io) or are built from Containerfiles in this repo.
  • Open-source Run:ai images and charts are public on NGC (nvcr.io/nvidia/runai-oss, helm.ngc.nvidia.com/nvidia/runai-oss); no NGC key is needed. The runai-installer binary is checked against the SHA-256 published with its GitHub release.
  • TLS for Run:ai uses a private CA created on the host (/var/lib/oif/pki, root only); only its public certificate is handed out for browsers to trust.

Host

  • Every host change is recorded and reversible (oif uninstall).
  • Passwordless sudo is added only for the installing user, only if needed by the Kubernetes tooling, and removed on uninstall.
  • GPUs not given to Kubernetes are isolated with vfio so workloads cannot reach them.

Workloads

  • Model-serving pods run as non-root where the engine allows it, with read-only root filesystems where possible, and request only the GPU resource they need.
  • Network policies limiting namespace-to-namespace traffic are planned for v0.2.

Reporting

Report vulnerabilities privately to the maintainer, not in public issues.

There aren't any published security advisories