Uh oh!
There was an error while loading. Please reload this page.
fix(ingestor): bump baseline image digest v0.3.2 → v0.3.6 (fixes a customer helm-upgrade regression) - #226
Closed
LukasWodka wants to merge 1 commit into
Closed
fix(ingestor): bump baseline image digest v0.3.2 → v0.3.6 (fixes a customer helm-upgrade regression)#226LukasWodka wants to merge 1 commit into
LukasWodka wants to merge 1 commit into
Conversation
The chart's baseline ingestor digest was stuck at v0.3.2 (2026-06-03), which predates the empty-column / NaN-tolerance ingest fixes (data-ingestors #150 / #167 / #168 + per-chunk type conversion #166). jobs-manager pins this digest, and a `helm upgrade` resets the live env to it — overwriting whatever image-refresh applied — so a customer `helm upgrade` silently reverts the ingestor to v0.3.2. This is the LMU regression: empty biomarker columns rejected as "non-numeric" again right after a helm upgrade (the CLI now correctly infers them FLOAT, but v0.3.2's float validator lacks the #150 NaN-tolerance). Bump to v0.3.6 (multi-arch index sha256:dfdaa7e6…, current stable with all the validator fixes; digest confirmed from the release-image CI run + the ingestor:0.3 float tag). Chart 1.5.0 -> 1.5.1. Follow-up (separate issue): `helm upgrade` should not clobber the image-refresh-managed digest — the imperative refresh and declarative chart fight, and upgrade wins. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This was referenced Jun 8, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Bump the chart's baseline ingestor digest from v0.3.2 → v0.3.6 (multi-arch index
sha256:dfdaa7e6…). The baseline had gone stale and was silently reverting customers to an old ingestor onhelm upgrade.Why (the a customer regression)
jobs-manager-deployment.yamlsetsINGESTOR_IMAGE_DIGESTfrom.Values.images.ingestor.digest. That baseline was pinned to v0.3.2 (2026-06-03) — which predates the empty-column / NaN ingest fixes (data-ingestors #150 float-NaN, #167 string-NaN, #168 date-NaN, #166 per-chunk conversion).The image-refresh CronJob had imperatively patched the live env (
kubectl set env) to the current0.3image, so ingestion worked. But a customerhelm upgradere-renders the deployment and resets the env back to the stale baseline digest, overwriting the refresh → the ingestor reverts to v0.3.2.Result at a customer: empty biomarker columns were rejected again — now as "non-numeric" (the CLI fix correctly types empty columns FLOAT, but v0.3.2's float validator lacks the #150 NaN-tolerance).
What
images.ingestor.digest→sha256:dfdaa7e633a8df0f46b403e9422eba5c16bdb7d9c39047e6d3738f9e38fbdba8(v0.3.6, current stable, all validator fixes).exporting manifest listindex from the v0.3.6 release-image CI run, andghcr.io/tracebloc/ingestor:0.3resolves to it — so it passes theingestor-multiarchCI gate.1.5.0 → 1.5.1.Follow-up (separate issue, filed)
The deeper bug:
helm upgrade(declarative) clobbers the image-refresh (imperativekubectl set env) — they fight and upgrade wins, so the digest silently regresses between releases. Needs a design fix (refresh writes release values, or the template doesn't reset a refresh-managed digest). Tracked separately.🤖 Generated with Claude Code