Compare commits

..

1 Commits

Author SHA1 Message Date
Barcode Betty 8ed8226e20 ci: never hard-fail deploy-dev/deploy-uat on infra-PR merge outcome (CAR-1216)
CI / test (pull_request) Successful in 13s
CI / lint (pull_request) Successful in 14s
CI / audit (pull_request) Failing after 13s
CI / build-and-push-receiptwitness (pull_request) Has been skipped
CI / build-and-push-api (pull_request) Has been skipped
CI / build-and-push-auth (pull_request) Has been skipped
CI / e2e (pull_request) Successful in 53s
CI / build-and-push (pull_request) Has been skipped
CI / deploy-dev (pull_request) Has been skipped
CI / deploy-uat (pull_request) Has been skipped
CI / lighthouse (pull_request) Failing after 1m24s
The CI bot cannot self-approve cartsnitch/infra main's required human
review, so the in-job auto-merge attempt is structurally impossible
and would always fail with SOME message (checks-pending, then
approvals, then transient errors). Special-casing individual error
strings is fragile — CAR-1212's 'Does not have enough approvals' branch
proved it (deploy-dev run 4999 hit the final else-branch instead).

Make job success depend only on the infra PR being OPENED (and
cs_savannah review requested), NOT on the merge outcome. Replace both
the approvals-elif branch and the final else (exit 1) branch with a
single non-failing branch: emit a ::notice:: with the merge response
and exit 0. The ONLY remaining hard-fail is the empty PR_NUM check
(PR could not be created).

Same change applied symmetrically to deploy-uat.

Refs CAR-1195, CAR-1194, CAR-1212, CAR-1216.
2026-06-03 22:05:37 +00:00
11 changed files with 51 additions and 121 deletions
+33 -57
View File
@@ -72,12 +72,6 @@ jobs:
lighthouse:
runs-on: ubuntu-latest
needs: [test]
# CAR-1218: continue-on-error until the Gitea Actions act runner can
# reliably capture lhci's stdout (currently suppressed — lhci exits
# ~40ms after start with no log output). The job still runs and
# reports; failures are surfaced on the PR but no longer block it.
# Quality-gate assertions in lighthouserc.json are unchanged.
continue-on-error: true
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020
@@ -91,28 +85,14 @@ jobs:
npm install -g playwright
npx playwright install --with-deps chromium
- name: Start preview server
# CAR-1218: bind to 127.0.0.1 (IPv4) not localhost. The act runner
# resolves 'localhost' to ::1 (IPv6) and the preview server does not
# get a reachable IPv4 socket, so wait-on times out.
run: |
npx vite preview --host 127.0.0.1 --port 4173 &
npx wait-on http://127.0.0.1:4173/ --timeout 30000
npm run preview &
npx wait-on http://localhost:4173/ --timeout 30000
- name: Run Lighthouse CI
# CAR-1218: act_runner does not honor continue-on-error at the job level
# (job still posts 'failure' status). Apply at the step level so the
# commit status reflects success and the PR is unblocked. lhci output
# is captured to a file (act_runner suppresses stdout from lhci).
continue-on-error: true
run: |
{
CHROME_PATH=$(find /home/runner/.cache/ms-playwright -name chrome -type f 2>/dev/null | head -1)
npm install -g @lhci/cli
CHROME_PATH="$CHROME_PATH" lhci autorun --chrome-flags="--headless=new --no-sandbox --disable-gpu --disable-dev-shm-usage"
} > /tmp/lhci.log 2>&1 || true
echo '=== lhci log (cat /tmp/lhci.log) ==='
cat /tmp/lhci.log || echo 'no lhci log produced'
echo '=== end lhci log ==='
exit 0
CHROME_PATH=$(find /home/runner/.cache/ms-playwright -name chrome -type f 2>/dev/null | head -1)
npm install -g @lhci/cli
CHROME_PATH="$CHROME_PATH" lhci autorun --chrome-flags="--headless=new --no-sandbox --disable-gpu --disable-dev-shm-usage"
build-and-push:
runs-on: ubuntu-latest
@@ -484,7 +464,7 @@ jobs:
with:
repository: cartsnitch/infra
token: ${{ secrets.CI_GITEA_TOKEN }}
ref: ${{ github.ref == 'refs/heads/main' && 'main' || (github.ref == 'refs/heads/uat' && 'uat' || 'dev') }}
ref: main
path: infra
- name: Install kubectl
@@ -508,14 +488,14 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update frontend image tag
if: needs.build-and-push.result == 'success'
run: |
cd infra/apps/overlays/dev
kustomize edit set image ghcr.io/cartsnitch/app=git.farh.net/cartsnitch/cartsnitch:${{ steps.frontend_tag.outputs.tag }}
kustomize edit set image ghcr.io/cartsnitch/cartsnitch=git.farh.net/cartsnitch/cartsnitch:${{ steps.frontend_tag.outputs.tag }}
- name: Determine image tag for receiptwitness
id: receiptwitness_tag
@@ -523,7 +503,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-receiptwitness.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-receiptwitness.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update receiptwitness image tag
@@ -538,7 +518,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-api.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-api.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update api image tag
@@ -553,7 +533,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-auth.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-auth.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update auth image tag
@@ -597,16 +577,6 @@ jobs:
if [ "${REVIEW_HTTP}" -lt 200 ] || [ "${REVIEW_HTTP}" -ge 300 ]; then
echo "::notice::Failed to request reviewers for cartsnitch/infra PR #${PR_NUM} (HTTP ${REVIEW_HTTP}); continuing"
fi
# CAR-1216: the in-job merge attempt is a best-effort fast-path only.
# `cartsnitch/infra` main requires a human approving review (immutable
# branch protection); the CI bot (`CI_GITEA_TOKEN`) can never self-
# approve, so this merge call structurally cannot succeed in the
# general case. Any non-merged outcome (approvals pending, checks
# pending, any other Gitea message) is the GitOps approval gate, not
# a CI failure — the PR is already opened and `cs_savannah` is
# requested as reviewer above. Surface the response as a notice and
# exit success. The only hard-fail (`exit 1`) in this step remains
# the empty-`PR_NUM` check (PR could not be created at all).
MERGE_RESP=$(curl -sS -X POST \
-H "Authorization: token ${CI_GITEA_TOKEN}" \
-H "Content-Type: application/json" \
@@ -616,6 +586,14 @@ jobs:
if [ "$MERGED" = "true" ]; then
echo "PR #${PR_NUM} merged into cartsnitch/infra main"
else
# GitOps approval gate (CAR-1216): the CI bot cannot self-approve
# cartsnitch/infra main's required human review, so the in-job
# auto-merge is structurally impossible and will always fail with
# SOME message (checks-pending, then approvals, then transient
# errors). Special-casing individual messages is fragile. The PR
# is correctly opened, cs_savannah is requested as reviewer, and
# the CTO will backstop-merge — that is the success condition.
# Treat every merge-outcome failure as success (exit 0).
echo "::notice::infra PR #${PR_NUM} opened and awaiting CTO (cs_savannah) approve+merge — GitOps approval gate, not a failure: $MERGE_RESP"
exit 0
fi
@@ -630,7 +608,7 @@ jobs:
with:
repository: cartsnitch/infra
token: ${{ secrets.CI_GITEA_TOKEN }}
ref: ${{ github.ref == 'refs/heads/main' && 'main' || (github.ref == 'refs/heads/uat' && 'uat' || 'dev') }}
ref: main
path: infra
- name: Install kubectl
@@ -654,14 +632,14 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update frontend image tag
if: needs.build-and-push.result == 'success'
run: |
cd infra/apps/overlays/uat
kustomize edit set image ghcr.io/cartsnitch/app=git.farh.net/cartsnitch/cartsnitch:${{ steps.frontend_tag.outputs.tag }}
kustomize edit set image ghcr.io/cartsnitch/cartsnitch=git.farh.net/cartsnitch/cartsnitch:${{ steps.frontend_tag.outputs.tag }}
- name: Determine image tag for receiptwitness
id: receiptwitness_tag
@@ -669,7 +647,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-receiptwitness.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-receiptwitness.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update receiptwitness image tag
@@ -684,7 +662,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-api.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-api.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update api image tag
@@ -699,7 +677,7 @@ jobs:
if [ "${{ github.ref }}" == "refs/heads/main" ]; then
echo "tag=${{ needs.build-and-push-auth.outputs.calver_tag }}" >> "$GITHUB_OUTPUT"
else
echo "tag=sha-${GITHUB_SHA}" >> "$GITHUB_OUTPUT"
echo "tag=${{ needs.build-and-push-auth.outputs.sha_tag }}" >> "$GITHUB_OUTPUT"
fi
- name: Update auth image tag
@@ -743,16 +721,6 @@ jobs:
if [ "${REVIEW_HTTP}" -lt 200 ] || [ "${REVIEW_HTTP}" -ge 300 ]; then
echo "::notice::Failed to request reviewers for cartsnitch/infra PR #${PR_NUM} (HTTP ${REVIEW_HTTP}); continuing"
fi
# CAR-1216: the in-job merge attempt is a best-effort fast-path only.
# `cartsnitch/infra` main requires a human approving review (immutable
# branch protection); the CI bot (`CI_GITEA_TOKEN`) can never self-
# approve, so this merge call structurally cannot succeed in the
# general case. Any non-merged outcome (approvals pending, checks
# pending, any other Gitea message) is the GitOps approval gate, not
# a CI failure — the PR is already opened and `cs_savannah` is
# requested as reviewer above. Surface the response as a notice and
# exit success. The only hard-fail (`exit 1`) in this step remains
# the empty-`PR_NUM` check (PR could not be created at all).
MERGE_RESP=$(curl -sS -X POST \
-H "Authorization: token ${CI_GITEA_TOKEN}" \
-H "Content-Type: application/json" \
@@ -762,6 +730,14 @@ jobs:
if [ "$MERGED" = "true" ]; then
echo "PR #${PR_NUM} merged into cartsnitch/infra main"
else
# GitOps approval gate (CAR-1216): the CI bot cannot self-approve
# cartsnitch/infra main's required human review, so the in-job
# auto-merge is structurally impossible and will always fail with
# SOME message (checks-pending, then approvals, then transient
# errors). Special-casing individual messages is fragile. The PR
# is correctly opened, cs_savannah is requested as reviewer, and
# the CTO will backstop-merge — that is the success condition.
# Treat every merge-outcome failure as success (exit 0).
echo "::notice::infra PR #${PR_NUM} opened and awaiting CTO (cs_savannah) approve+merge — GitOps approval gate, not a failure: $MERGE_RESP"
exit 0
fi
-2
View File
@@ -1,2 +0,0 @@
# CAR-1374 verification no-op
2026-06-10T22:57:17Z CAR-1375 uat regression trigger
-2
View File
@@ -313,5 +313,3 @@ Secrets are managed via **Bitnami Sealed Secrets**. No plain Kubernetes secrets
## License
MIT © 2025 CartSnitch
<!-- CAR-1371 verification: trigger deploy-dev to confirm --arg base dev -->
+2 -1
View File
@@ -31,6 +31,7 @@ def run_migrations_offline() -> None:
target_metadata=target_metadata,
literal_binds=True,
dialect_opts={"paramstyle": "named"},
version_table_column_width=128,
)
with context.begin_transaction():
context.run_migrations()
@@ -44,7 +45,7 @@ def run_migrations_online() -> None:
poolclass=pool.NullPool,
)
with connectable.connect() as connection:
context.configure(connection=connection, target_metadata=target_metadata)
context.configure(connection=connection, target_metadata=target_metadata, version_table_column_width=128)
with context.begin_transaction():
context.run_migrations()
# Create any tables defined in models but not yet created by migrations.
@@ -33,15 +33,6 @@ def _is_fernet_token(value: str) -> bool:
def upgrade() -> None:
# Alembic hardcodes alembic_version.version_num to VARCHAR(32)
# (DefaultImpl.version_table_impl) and exposes no option to widen it
# (version_table_column_width is NOT a real kwarg — it is silently ignored).
# Our descriptive revision ids exceed 32 chars (e.g.
# 003_make_users_hashed_password_nullable = 39), so widen the column as the
# very first migration statement, before any early-return path below.
# Idempotent: a no-op when already wider (e.g. pre-created by the CAR-1298 Job).
op.execute("ALTER TABLE alembic_version ALTER COLUMN version_num TYPE VARCHAR(128)")
conn = op.get_bind()
inspector = sa.inspect(conn)
+2 -2
View File
@@ -1,4 +1,4 @@
FROM node:22-alpine@sha256:8ea2348b068a9544dae7317b4f3aafcdc032df1647bb7d768a05a5cad1a7683f AS builder
FROM node:22-alpine AS builder
RUN apk update && apk upgrade --no-cache
WORKDIR /app
COPY package.json package-lock.json* ./
@@ -7,7 +7,7 @@ COPY tsconfig.json ./
COPY src/ src/
RUN npm run build
FROM node:22-alpine@sha256:8ea2348b068a9544dae7317b4f3aafcdc032df1647bb7d768a05a5cad1a7683f
FROM node:22-alpine
RUN apk update && apk upgrade --no-cache
WORKDIR /app
ENV NODE_ENV=production
+4 -23
View File
@@ -19,18 +19,9 @@ describe('Auth health endpoint', () => {
}
res.writeHead(200, { 'Content-Type': 'application/json' });
res.end(JSON.stringify({ status: 'ok', db: 'reachable' }));
} catch (err) {
// Mirror src/index.ts: log the error and include the message in the
// response body so /health 503s are diagnosable from pod logs.
console.error(
'[auth /health] DB probe failed:',
err instanceof Error ? `${err.name}: ${err.message}` : err,
);
const detail = err instanceof Error ? err.message : 'unknown error';
} catch {
res.writeHead(503, { 'Content-Type': 'application/json' });
res.end(
JSON.stringify({ status: 'error', db: 'unreachable', error: detail }),
);
res.end(JSON.stringify({ status: 'error', db: 'unreachable' }));
}
return;
}
@@ -85,10 +76,7 @@ describe('Auth health endpoint', () => {
close();
equal(status, 503);
const parsed = JSON.parse(body);
equal(parsed.status, 'error');
equal(parsed.db, 'unreachable');
equal(parsed.error, 'connection refused');
equal(body, '{"status":"error","db":"unreachable"}');
});
it('returns 503 with db=unreachable when query times out', async () => {
@@ -107,14 +95,7 @@ describe('Auth health endpoint', () => {
close();
equal(status, 503);
const parsed = JSON.parse(body);
equal(parsed.status, 'error');
equal(parsed.db, 'unreachable');
// The query promise rejects with a synthetic 'timeout' error; the
// Promise.race wrapper also rejects with 'DB timeout'. The body should
// surface whichever error was thrown — accept either to stay robust.
equal(typeof parsed.error, 'string');
equal(parsed.error.length > 0, true);
equal(body, '{"status":"error","db":"unreachable"}');
});
it('returns a terminal response for unknown paths (no hang)', async () => {
+2 -12
View File
@@ -21,19 +21,9 @@ const server = createServer(async (req, res) => {
}
res.writeHead(200, { "Content-Type": "application/json" });
res.end(JSON.stringify({ status: "ok", db: "reachable" }));
} catch (err) {
// Log the actual error so /health 503s are diagnosable from pod logs
// (CAR-1276: UAT auth was crashlooping with no log output beyond the
// initial "listening on port 3001" line because this catch was empty).
console.error(
"[auth /health] DB probe failed:",
err instanceof Error ? `${err.name}: ${err.message}` : err,
);
const detail = err instanceof Error ? err.message : "unknown error";
} catch {
res.writeHead(503, { "Content-Type": "application/json" });
res.end(
JSON.stringify({ status: "error", db: "unreachable", error: detail }),
);
res.end(JSON.stringify({ status: "error", db: "unreachable" }));
}
return;
}
@@ -18,11 +18,6 @@ depends_on: str | Sequence[str] | None = None
def upgrade() -> None:
# Same VARCHAR(32) alembic_version limitation as the api migrations; the
# common 002 revision id is 46 chars. Widen first so a fresh-DB upgrade can
# stamp it. Idempotent.
op.execute("ALTER TABLE alembic_version ALTER COLUMN version_num TYPE VARCHAR(128)")
op.add_column("users", sa.Column("email_inbound_token", sa.String(22), nullable=True))
op.create_unique_constraint("uq_users_email_inbound_token", "users", ["email_inbound_token"])
+1 -1
View File
@@ -2,7 +2,7 @@
"ci": {
"collect": {
"staticDistDir": "./dist",
"url": ["http://127.0.0.1:4173/"],
"url": ["http://localhost:4173/"],
"numberOfRuns": 1,
"settings": {
"chromeFlags": ["--headless=new", "--no-sandbox", "--disable-gpu", "--disable-dev-shm-usage"],
+7 -7
View File
@@ -8305,9 +8305,9 @@
}
},
"node_modules/react-router": {
"version": "7.16.0",
"resolved": "https://registry.npmjs.org/react-router/-/react-router-7.16.0.tgz",
"integrity": "sha512-wArC8lVyJb3+jM9OpDyW6hLCizACWkvQR/sSGqSs+o5uEXEtGlqdZ4v8hENR3Jad6i+LRkK93q/+bQAcvl6V1A==",
"version": "7.14.0",
"resolved": "https://registry.npmjs.org/react-router/-/react-router-7.14.0.tgz",
"integrity": "sha512-m/xR9N4LQLmAS0ZhkY2nkPA1N7gQ5TUVa5n8TgANuDTARbn1gt+zLPXEm7W0XDTbrQ2AJSJKhoa6yx1D8BcpxQ==",
"license": "MIT",
"dependencies": {
"cookie": "^1.0.1",
@@ -8327,12 +8327,12 @@
}
},
"node_modules/react-router-dom": {
"version": "7.16.0",
"resolved": "https://registry.npmjs.org/react-router-dom/-/react-router-dom-7.16.0.tgz",
"integrity": "sha512-kMUAbimWB5FVbF4Bce4bJsiKJWLIUHq/mEG8+CFDnCSgltptBiG5nguducmsJeGKytlCvQud9Qhzpn49iduTlA==",
"version": "7.14.0",
"resolved": "https://registry.npmjs.org/react-router-dom/-/react-router-dom-7.14.0.tgz",
"integrity": "sha512-2G3ajSVSZMEtmTjIklRWlNvo8wICEpLihfD/0YMDxbWK2UyP5EGfnoIn9AIQGnF3G/FX0MRbHXdFcD+rL1ZreQ==",
"license": "MIT",
"dependencies": {
"react-router": "7.16.0"
"react-router": "7.14.0"
},
"engines": {
"node": ">=20.0.0"