wbsong111
2af7fcb43d
manifests/helm 카탈로그 전 이미지의 SBOM(CycloneDX)과 취약점 리포트를 생성하는 파이프라인 추가. 컨테이너 내부 trivy/helm 바이너리 직접 호출. - doc/scripts/extract-helm-images.sh: helm template 렌더로 이미지 인벤토리 추출 - doc/scripts/generate-sbom.sh: trivy image --format cyclonedx (이미지 1회 pull) - doc/scripts/scan-sbom.sh: trivy sbom 로 SBOM 기반 오프라인 취약점 스캔 - .github/workflows/sbom.yml: 스케줄(전체)+PR(증분)+수동, 산출물은 아티팩트 - doc/sbom-pipeline.md: 파이프라인 설계·운영 문서(단일) - latest/무태그 digest 고정: mlflow postgresql, vllm-openai(v0.24.0) - .gitignore: 파이프라인 산출물 제외 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
vLLM Production Stack helm chart
This helm chart lets users deploy multiple serving engines and a router into the Kubernetes cluster.
Key features
- Support running multiple serving engines with multiple different models
- Load the model weights directly from the existing PersistentVolumes
Prerequisites
- A running Kubernetes cluster with GPU. (You can set it up through
minikube: https://minikube.sigs.k8s.io/docs/tutorials/nvidia/) - Helm
Install the helm chart
helm install llmstack . -f values-example.yaml
Uninstall the deployment
run helm uninstall llmstack
Configure the deployments
See helm/values.yaml for mode details.