Skip to content

Omnia Documentation

Omnia version Downloads Last Commit Contributors Forks License

Omnia is a containerized, open-source deployment toolkit designed to automate the setup and management of high-performance computing (HPC) environments on Linux-based servers. It leverages Ansible playbooks to streamline:

  • Operating system provisioning
  • Driver installation and configuration
  • Deployment of workload schedulers such as Slurm and Kubernetes
  • Installation of optimization libraries, machine learning frameworks, and AI models
  • Management of compute, storage, and networking resources

Omnia simplifies infrastructure deployment in complex environments, enabling faster setup and consistent configuration across systems.

The project is hosted on GitHub, where you can:

  • Access the source code
  • Report issues
  • Ask questions
  • Contribute to development

How This Documentation is Organized

  • Overview

    Architecture, components, network topologies, and design concepts. Start here if you are new to Omnia.

  • Get Started

    End-to-end tutorials that take you from a bare set of PowerEdge servers to a fully operational cluster. Choose from Slurm-only, full deployment, Kubernetes + telemetry, or BuildStreaM paths.

  • How-to Guides

    Task-oriented procedures for provisioning, configuring Slurm, Kubernetes, storage, networking, authentication, telemetry, and BuildStreaM.

  • Reference

    Configuration parameters, support matrices, playbook references, API documentation, and network port listings.

  • Operations & Maintenance

    Day-2 operations: adding and removing nodes, re-provisioning, upgrading and rolling back Omnia versions, OIM cleanup, log management, security hardening, and best practices.

  • Troubleshooting

    Symptom-driven guides for diagnosing and resolving issues with provisioning, Slurm, Kubernetes, telemetry, authentication, and more.

Resource Description
Slurm Quickstart Fastest path to a working Slurm cluster (~2 hours, 4 nodes).
Full Deployment Production deployment with Slurm, Kubernetes, telemetry, and LDAP.
Servers Supported OS versions, hardware, firmware, and software combinations.
Provision Config Complete reference for all Omnia input configuration files.

Licensing

Omnia is made available under the Apache 2.0 license.

Note

Omnia playbooks are licensed under the Apache 2.0 license. Once an end-user initiates Omnia, that end-user will deploy other open-source and/or third-party software that is licensed separately by their respective developer communities and/or third parties. For a comprehensive list of software and their licenses, click here. Dell (or any other contributors) shall have no liability regarding (and no responsibility to provide support for) an end-user's use of any open-source and/or third-party software and Omnia users are solely responsible for ensuring that they are complying with all such licenses. Omnia is provided "as is" without any warranty, express or implied. Dell (or any other contributors) shall have no liability for any direct, indirect, incidental, punitive, special, or consequential damages for an end-user's use of Omnia.

Previous Versions

For a better understanding of what Omnia does, check out the following:

Note

Upgrade from Omnia 1.x to 2.x is not supported due to architectural changes.

Omnia Community Members

Dell Technologies Intel University of Pisa Community Member VIZIAS Community Member Liqid

If you have any feedback about Omnia documentation, please reach out at omnia.readme@dell.com.