Deploy and manage Slurm-on-Kubernetes clusters on Nebius via declarative YAML.
npx skills add https://github.com/nebius/nebius-physical-ai --skill soperator
Please help me install this Agent Skill. Skill: soperator Source: https://github.com/nebius/nebius-physical-ai/tree/main/skills/tools/soperator Command: npx skills add https://github.com/nebius/nebius-physical-ai --skill soperator
This skill simplifies the complex deployment and lifecycle management of Slurm-on-Kubernetes clusters on Nebius, removing the need for manual Terraform configuration and infrastructure tuning.
Use the soperator skill to deploy a new Slurm cluster by running the command npa soperator deploy with your cluster specification file.
High-intent search queries and answers about installing and using this skill.
To deploy a Slurm cluster on Kubernetes, you use declarative YAML specifications with the soperator skill. It automates the Terraform configuration and infrastructure tuning required for production-ready Slurm-on-Kubernetes clusters on Nebius.
Yes, you can configure multi-preset worker pools optimized for GPU workloads. The skill supports node-local Docker and Enroot image caching to enhance performance for high-performance computing tasks.
Running Slurm on Nebius requires Terraform integration and specific compute and GPU quota preflight checks. You must ensure your infrastructure meets these resource quotas before deploying the cluster.
Automated cluster management handles status monitoring, post-deployment health fixes, and resource cleanup. This operational lifecycle automation removes the need for manual infrastructure tuning after your Slurm cluster is deployed.
Using declarative YAML for Slurm cluster deployment simplifies complex lifecycle management on Nebius. It removes manual Terraform configuration, allowing you to define production-ready clusters using the npa.soperator specification.
attentiondotnetValidate Azure configurations and infrastructure for deployment readiness.
briannadoubtConfigure structured logs, metrics, and traces for Trebuchet actors with CloudWatch export.
tomzConfigure CoreWeave GPU clusters with node selectors, MIG profiles, and NCCL tuning.
luiscamaralAutomate L0 infrastructure provisioning with Talos Linux and Proxmox Terraform deployments.
