Pin the Dockerfile clones on the production branch - #19
Open
dionmcm wants to merge 1 commit into
Open
Conversation
Both clones - snomed-drools-rules and snomed-release-validation-assertions - were unpinned `git clone`, so the image contents depend on the day it was built. PR #14 fixed this on the catch-up line (d5cb512) but never reached groups-api-fix, which is the branch production builds from. Demonstrated today while building an image to verify two unrelated fixes: a rebuild picked up assertions HEAD 0160dd2e, four files had been removed upstream since fad36466, and RVF died during startup - FileNotFoundException: ./snomed-release-validation-assertions/scripts/ release-type/release-type-delta-previous-snapshot-validation-inferred- relationship_EDITION.sql rvf-gate:local (last known good) fad36466 140 files in scripts/release-type rvf-gate:armE (fresh, unpinned) 0160dd2e 136 files - the reference is gone This is a startup failure, so the nightly produces no report at all rather than a red one. It is latent until something forces a rebuild - which is exactly what shipping the schema-mapping and column-count fixes requires. Pin first, then rebuild. fad36466 is chosen because it is what the working image carries and what every green nightly to date has used, not because it is current. Bumping it requires re-checking that every sqlFile reference in testscripts/manifest.xml resolves. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018aYBXcyiozxQQGhzzhS7KG
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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Merge this before anything that rebuilds the image. Right now a rebuild
produces an image that does not boot.
Both clones —
snomed-drools-rulesandsnomed-release-validation-assertions—are unpinned
git clone, so what the image contains depends on the day it wasbuilt. PR #14 fixed this on the catch-up line (
d5cb5129) but never reachedgroups-api-fix, which is the branch production builds from.Demonstrated today
While building an image to verify two unrelated fixes, the rebuild picked up
assertions HEAD
0160dd2e. Four files had been removed upstream sincefad36466, and RVF died during startup:scripts/release-typefad364660160dd2eThis is a startup failure, so the nightly produces no report at all rather
than a red one — the same shape as the 2026-08-07 outage, where a two-year-old
image was rebuilt and broke production for the same reason.
Why it matters now
The defect is latent until something forces a rebuild. Two fixes in flight do
exactly that: PR #18 (trailing empty column) and the schema-mapping fix. Either
one, merged and built before this, ships a nightly that cannot start.
On the chosen ref
fad36466is what the last known-good image carries and what every green nightlyto date has used. It is not the newest commit, and that is deliberate — bumping it
requires re-checking that every
sqlFilereference intestscripts/manifest.xmlstill resolves, since a single missing one stops RVF booting. Pinning first makes
that bump a reviewable change rather than something that happens by calendar.
No behaviour change: the pinned refs are what a working image already contains.