Deploy and manage distributed Ray clusters on Kubernetes with the KubeRay operator.
npx skills add https://github.com/tylertitsworth/skills --skill kuberay
Please help me install this Agent Skill. Skill: kuberay Source: https://github.com/tylertitsworth/skills/tree/main/kuberay Command: npx skills add https://github.com/tylertitsworth/skills --skill kuberay
Simplifies deploying and orchestrating distributed Ray workloads on Kubernetes by providing a dedicated operator and CRDs.
Install the KubeRay operator and apply a RayCluster manifest to launch a head and workers.
High-intent search queries and answers about installing and using this skill.
Deploy Ray clusters on Kubernetes by using the KubeRay operator to apply CRD manifests for RayCluster resources, launching a head node and worker pods automatically. Helm-based installation sets up the operator, then you apply your RayCluster manifest to start distributed workloads.
RayServe runs on Kubernetes by creating RayService custom resources managed by the KubeRay operator. This deploys and manages RayServe workloads natively, integrating autoscaling and TLS authentication for scalable model serving.
KubeRay supports autoscaling for Ray worker nodes within Kubernetes environments. The operator dynamically scales worker pods based on workload demand, managed via CRD-driven resource configurations.
Enable GCS fault tolerance by configuring the KubeRay operator settings in your RayCluster manifest. This ensures distributed Ray workloads maintain state and recover from head node failures across the Kubernetes cluster.
Running Ray workloads requires Kubernetes access with proper RBAC permissions to manage RayCluster, RayJob, and RayService CRDs. The KubeRay operator needs permissions to create and manage pods, services, and custom resources across multi-tenant environments.
Ray jobs run in multi-tenant Kubernetes environments using KubeRay CRDs for tenant isolation. The operator manages RayJob resources with queue-based scheduling integration via Kueue, ensuring robust workload isolation and observability.
dieu-donneeDeploy Kubernetes workloads with declarative manifests, RBAC, and network policies.
microsoftAutomate Azure observability workflows for monitoring and diagnosing cloud workloads.
haxlysConfigure mutual TLS policies for Istio, Linkerd, SPIRE, and cert-manager deployments.
ovachieverOrchestrate distributed training across clusters with Ray Train and Ray Tune.
