mirror of
https://github.com/Sea-Haven-Industries/open-swe.git
synced 2026-09-30 06:53:14 +00:00
3224d7abf4
Some checks are pending
Build & publish app artifacts / Publish + deploy (dev) (push) Waiting to run
Build & publish app artifacts / Publish + deploy (prod) (push) Waiting to run
Infra CD / Infra CI (pre-deploy) (push) Waiting to run
Infra CD / Deploy open-swe-dev (push) Blocked by required conditions
Infra CD / Deploy open-swe-prod (push) Blocked by required conditions
Agent CI / Agent lint (push) Waiting to run
Agent CI / Agent format check (push) Waiting to run
Agent CI / Agent unit tests (push) Waiting to run
Agent CI / Playwright E2E (push) Waiting to run
T20 CD safety nets. Two gaps closed before the first real prod deploy: 1. Promotion gate. promote_dev_to_prod.yml previously fast-forwarded dev→main unconditionally. It now hard-gates on check-dev-green.sh: every check-run on the dev HEAD must be completed+passing AND the Agent CI suite (lint/format/unit/E2E) must be present+success, or the promotion blocks (fails safe on a missing/renamed check). The promote run excludes its OWN check-run by run-id (unforgeable), never by the mutable name "promote", so a colliding red check cannot hide. Fields are read with a 0x1F separator so an empty conclusion (every in_progress check) cannot shift columns. ci.yml now also runs on push:dev so dev HEAD actually carries that signal (a PR check alone can be admin-merged past). 2. Rollback + last-good. publish-and-deploy.sh advances releases/last-good/ only after a successful roll (deploy.sh gates on `systemctl is-active`), and makes releases/latest/ transactional — reverting to the prior release if the roll fails so a replaced box never self-deploys a broken release. New rollback.yml + rollback.sh re-point latest at last-good (or an explicit sha) and re-fire the deploy; prod is gated by the `prod` Environment approval, same as a deploy. The shared fire/wait/aggregate-gate logic is factored into roll-box.sh (used by both forward and backward rolls). Least-privilege: drop the unused s3:DeleteObject from the app deploy role — publish/rollback/deploy only Get+Put (S3-to-S3 copy), and the rollback fallback now depends on immutable release history staying intact. Lifecycle expiry (not CI) handles old-version cleanup. Gate logic unit-tested (7 cases + jq round-trip). IAM change + release-safety control cross-reviewed by GPT-4.1: APPROVE, no blocks. Claude-Session: https://claude.ai/code/session_01DMhLf4G5V8MStJQyAW95hi
85 lines
3.3 KiB
Bash
Executable file
85 lines
3.3 KiB
Bash
Executable file
#!/usr/bin/env bash
|
|
# Promotion gate: refuse to promote a dev HEAD that is not fully green.
|
|
#
|
|
# Reads check-runs on stdin — one
|
|
# name<US>status<US>conclusion<US>details_url
|
|
# per line, fields separated by ASCII Unit Separator (0x1F) — so it is unit-testable
|
|
# WITHOUT GitHub. promote_dev_to_prod.yml pipes the live `gh api .../check-runs`
|
|
# output in. 0x1F (not TAB) is used deliberately: TAB is IFS-whitespace, so an empty
|
|
# conclusion (every in_progress check has a null conclusion) would collapse and shift
|
|
# the columns — which would make the promote run fail to exclude itself. 0x1F is
|
|
# non-whitespace, so `read` preserves empty fields and the columns stay aligned.
|
|
#
|
|
# A dev HEAD is promotable ONLY when BOTH hold:
|
|
# 1. every present check-run is completed with a passing conclusion
|
|
# (success/neutral/skipped); any pending/failed/cancelled/timed_out check BLOCKS.
|
|
# 2. every check named in REQUIRED_CHECKS is present AND concluded "success".
|
|
# This positive allow-list is what stops a partial-signal promotion — e.g. a
|
|
# commit pushed with [skip ci] (no CI check-runs) that still carries one
|
|
# unrelated green check, or a required check silently renamed/dropped. The gate
|
|
# FAILS SAFE: a missing required check blocks rather than promotes.
|
|
#
|
|
# Self-exclusion: the promotion workflow's OWN in-progress check-run is dropped by
|
|
# EXCLUDE_RUN_ID (its github.run_id, matched in the check-run details_url) — an
|
|
# unforgeable identity, NOT a mutable check name. A check that merely happens to be
|
|
# named "promote" can no longer hide a real failing/required check.
|
|
set -euo pipefail
|
|
|
|
EXCLUDE_RUN_ID="${EXCLUDE_RUN_ID:-}"
|
|
# Mandatory checks (one per line). Defaults to the Agent CI suite, which runs on
|
|
# every push to dev (see ci.yml). Keep in sync with those job names; if a name
|
|
# drifts the gate blocks (fails safe) until the list is updated.
|
|
REQUIRED_CHECKS="${REQUIRED_CHECKS:-Agent lint
|
|
Agent format check
|
|
Agent unit tests
|
|
Playwright E2E}"
|
|
|
|
declare -A GREEN
|
|
seen=0
|
|
bad=0
|
|
while IFS=$'\037' read -r name status conclusion details_url; do
|
|
[ -n "${name:-}" ] || continue
|
|
# drop the promotion run's own check-run by run id (never by name).
|
|
if [ -n "${EXCLUDE_RUN_ID}" ] && \
|
|
[ "${details_url}" != "${details_url#*/runs/${EXCLUDE_RUN_ID}/}" ]; then
|
|
continue
|
|
fi
|
|
seen=$((seen + 1))
|
|
if [ "${status}" != "completed" ]; then
|
|
echo "BLOCK: check '${name}' is '${status}' (not completed)" >&2
|
|
bad=1
|
|
continue
|
|
fi
|
|
case "${conclusion}" in
|
|
success)
|
|
GREEN["${name}"]=1
|
|
echo "ok: ${name} (success)" ;;
|
|
neutral | skipped)
|
|
echo "ok: ${name} (${conclusion})" ;;
|
|
*)
|
|
echo "BLOCK: check '${name}' concluded '${conclusion:-<none>}'" >&2
|
|
bad=1 ;;
|
|
esac
|
|
done
|
|
|
|
if [ "${seen}" -eq 0 ]; then
|
|
echo "BLOCK: no check-runs found for this commit — refusing to promote an unverified dev HEAD" >&2
|
|
exit 1
|
|
fi
|
|
|
|
missing=0
|
|
while IFS= read -r req; do
|
|
[ -n "${req}" ] || continue
|
|
if [ -z "${GREEN[${req}]:-}" ]; then
|
|
echo "BLOCK: required check '${req}' is missing or not successful on dev HEAD" >&2
|
|
missing=1
|
|
fi
|
|
done <<EOF
|
|
${REQUIRED_CHECKS}
|
|
EOF
|
|
|
|
if [ "${bad}" -ne 0 ] || [ "${missing}" -ne 0 ]; then
|
|
echo "BLOCK: dev HEAD is not promotable (see reasons above)" >&2
|
|
exit 1
|
|
fi
|
|
echo "PASS: ${seen} check(s) present, all green; all required checks present + successful."
|