gabriel / musehub public
Closed #165 Enhancement production-readiness
filed by gabriel human · 68 days ago

Production readiness — Section 17: Testing Before Launch

0 Anchors
— Blast radius
— Churn 30d
0 Proposals

Full detail

Verified the automated test suite exists and collects cleanly: 7,028 tests, zero collection errors. Actually running the full suite for pass/fail status is still open. Several items (backup/restore testing, canonical-redirect testing) cannot pass until their prerequisite sections (8, 6) are built.

Also confirmed failing today: "verify neither maintainer is a single point of operational failure" — currently fails due to the shared deploy credential and Aaron's missing AWS identity (Section 1).

◎ Activity1
●
gabriel opened this issue 68 days ago
○
gabriel 37 days ago

Ran the actual test suite for the first time (previously only confirmed it collects): 7,028 tests, 7,003 passed / 15 failed / 24 skipped. Root-caused and fixed 9 of 15 failures — a real bug in a test helper (_attach_repo_to_domain set the wrong DB column, domain_id instead of marketplace_domain_id — application code was correct, only the test fixture was wrong) plus a clean regression (Dockerfile's PLAYWRIGHT_BROWSERS_PATH never added to the approved safe-env-var test list). Remaining 5 failures are a test-environment gap (missing blob storage config), not app bugs.

Also confirmed live, externally:

  • Canonical redirects (www→apex, http→https) — both 301, verified via curl
  • TLS from an external network — valid cert, correct CN, via openssl
  • Maintainer-SPOF test now passes — was failing per the original doc; resolved via #150 (Aaron has independent SSO identity, deploys staging/production under his own identity)
  • Rollback already tested for real (#159)

Remaining real gaps (integration/load/soak testing, tabletop exercise, break-glass test, the OpenGraph test fix) split to #179.

Closing.

✓
closed this issue 37 days ago