What is your question?
I've been contributing to the repo and submitted PR #13 (ARCHITECTURE.md). I'm also working on a Ray Serve + KubeRay Helm chart example based on the community invitation in the launch announcement.
To test any of this end-to-end, community contributors need access to the NGC nvcf-onprem org to pull the control plane images during helmfile sync. Without it, the cluster bootstrap works fine but the actual NVCF services cannot be deployed.
Is there a formal path for external contributors to get NGC nvcf-onprem access for testing purposes? Or is the expectation that contributors submit PRs without local end-to-end validation and the team tests internally when carrying them through the pipeline?
My NGC details for reference:
Happy to share any write-ups or test results with the team once I have access.
Related: #12 (full OSS build would eventually eliminate this dependency entirely for the community)
What is your question?
I've been contributing to the repo and submitted PR #13 (ARCHITECTURE.md). I'm also working on a Ray Serve + KubeRay Helm chart example based on the community invitation in the launch announcement.
To test any of this end-to-end, community contributors need access to the NGC
nvcf-onpremorg to pull the control plane images duringhelmfile sync. Without it, the cluster bootstrap works fine but the actual NVCF services cannot be deployed.Is there a formal path for external contributors to get NGC
nvcf-onpremaccess for testing purposes? Or is the expectation that contributors submit PRs without local end-to-end validation and the team tests internally when carrying them through the pipeline?My NGC details for reference:
Happy to share any write-ups or test results with the team once I have access.
Related: #12 (full OSS build would eventually eliminate this dependency entirely for the community)