NVIDIA / NVIDIA/nvcf

release: helmfile layer batching makes needs: edges unsafe to add

Open
#1,981 0 comments 0 reactions 1 assignee View on GitHub

@kristinapathak is already working on this.

Since Sep 18, 2026.

ci
Dominant language
Go
Stars
218
Forks
72
Avg merge
1d 12h
Merged PRs (30d)
427

Description

deploy/stacks/self-managed/helmfile.d/02-core.yaml.gotmpl documents a needs: edge that had to be removed. Read the comment on the admin-issuer-proxy release for the full root cause: helmfile 1.1.9 processes releases layer-by-layer rather than strictly DAG-aware, so adding a needs: put the release in DAG layer 2 where it waited on all layer-1 peers rather than its actual prerequisite. When an unrelated layer-1 release hung, layer-2 releases were silently skipped — helmfile list reported them installed while helm list -A showed no release.

Meanwhile Makefile.dist pins helmfile to 1.1.x because 1.2.0+ processes helmfile.d/ in parallel and breaks the implicit alphabetical stage ordering the stack currently depends on.

So the stack is wedged: the version that preserves today's implicit ordering is the version that makes explicit ordering hazardous. Resolving this is a prerequisite for declaring the real dependency graph.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.