NVIDIA / NVIDIA/cluster-readiness-engine
[Feature]: Air-gapped install
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 55
- Forks
- 19
- Avg merge
- 1d 5h
- Merged PRs (30d)
- 107
Description
Prerequisites
- I searched existing issues
Code of Conduct
- I agree to follow the project's Code of Conduct
Feature Summary
Install, run, and remove NVCRE with no internet access.
Problem/Use Case
A significant share of GPU clusters run disconnected by policy. Install is Helm-based today so images are mirrorable, but the test catalog has no defined distribution model — which means there's no supported way to get tests and thresholds into an air-gapped environment, and no way to update them.
Proposed Solution
Publish a complete image manifest operators can mirror to a private registry in one step, and make the catalog resolvable from a mirrored OCI registry via a documented override. Validate the whole path — install, certify, tear down — in an environment with egress denied.
Component
Helm / Deployment
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by tracing the Helm-based installation and the test catalog distribution model, then identify how mirrored OCI registry settings enter deployment. Done means operators can mirror the complete image manifest, resolve the catalog from the private registry, and validate install, certification, and teardown with egress denied.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go, helm, kubernetes
- Domain
- devops, infrastructure
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100