Files
edr-platform/docker-compose.it.yaml
Nathnael 415ae52143 test(integration): boot freight API in-process across parallel shards
The suite drove a containerized freight API, so every code change needed an
image rebuild before a test could see it, and there was no way to attach a
debugger. Files also ran strictly in sequence against one shared database,
which is the root of the warm-stack gotchas the README documents: stowaway
paid bookings climbing back aboard, a short consist on the fifth file.

The freight app now boots inside each vitest worker from dist/, and each
worker owns a whole shard of the topology - its own database, payment API,
gateway mock and broker vhost - so nothing mutable is shared and files run
in parallel. Full suite drops from roughly 20 minutes to 196s at 4 shards.

- main.ts exports createFreightApp() so the harness applies the same prefix,
  pipes, filters and interceptors as production instead of replaying them by
  hand; self-start is guarded by require.main so the Dockerfile CMD still boots
- booking-window tick cadence is env-driven (BOOKING_WINDOW_TICK_CRON), */1 in
  the suite, */10 unchanged in production
- prepare-shards.mjs seeds a template database (boot seeders, then the SQL
  fixtures that depend on them) and clones it per shard; it.mjs re-clones on
  every run, so each run is hermetic
- gateway mock and payment API are generated per shard: the mock keeps modes
  and orders process-global and 20 of 25 specs reset it in beforeAll, and the
  inbound CBE bill query has to reach one specific shard's app
- poll() samples every 250ms instead of 2000ms, keeping the caller's deadline
- authz.it.ts seeds its own invoice; it previously read another spec's leftover
  and returned early, which silently passed on a pristine database

Known: an unlocked MAX(sequence_no)+1 in train-scheduling.service.ts races
under concurrent allocation and leaves a short consist, so 1-3 specs fail
intermittently. Pre-existing and reproduces at the production tick cadence.
2026-08-04 12:43:25 +00:00

61 lines
2.9 KiB
YAML

# EDR Freight — API integration stack (headless, sharded).
#
# Overlay on docker-compose.e2e.yaml, plus a GENERATED third file holding the
# per-shard services (integration/.it-shards.yaml, written by gen-shards.mjs).
#
# The freight API is NOT a container here — it boots inside each vitest worker,
# on the host, so a code change needs no image rebuild and a breakpoint works.
# Each worker is a full shard of the topology; nothing mutable is shared:
#
# shard i:
# freight app (in vitest worker, host :3111+i)
# │ ▲
# │ └──── HTTP ─── payment-api-it-{i} :3131+i (inbound CBE bill query,
# │ │ via host.docker.internal)
# └── HTTP ──────────────► │ ──HTTP──> gateway-mock-it-{i} :4600+i
# ▲ │
# └──── RabbitMQ vhost payment_s{i} ──┘ (outbox → payment.events → consumer)
#
# Shared by every shard: postgres (one container, one database per shard cloned
# from a seeded template), rabbitmq (one container, one vhost per shard), minio,
# and the one-shot migration.
#
# Its own compose project (`name:` below overrides the base) and its own host
# ports, so it can run side by side with the Cypress e2e stack.
#
# node integration/scripts/it.mjs up|test|down|logs
#
# Never start it with plain `docker compose -f docker-compose.it.yaml` — it is
# an OVERLAY and needs the base file first:
# docker compose -f docker-compose.e2e.yaml -f docker-compose.it.yaml \
# -f integration/.it-shards.yaml ...
name: edr-freight-it
services:
# Outbox transport. The payment API publishes payment.succeeded/failed here
# and freight consumes it — the production path. Copied from the passenger
# harness (e2e/docker-compose.yml). One broker, one vhost per shard
# (payment_s0, payment_s1, …) created by it.mjs — a shared vhost would let one
# shard's freight consumer eat another shard's settlement event.
rabbitmq-it:
image: rabbitmq:3-management
environment:
RABBITMQ_DEFAULT_USER: edr
RABBITMQ_DEFAULT_PASS: edr_secret
RABBITMQ_DEFAULT_VHOST: payment
ports:
- "${IT_RABBIT_PORT:-5772}:5672"
- "${IT_RABBIT_UI_PORT:-15772}:15672"
healthcheck:
# check_running, NOT ping: ping only proves the Erlang node answers, and
# it.mjs runs `rabbitmqctl add_vhost` the moment this goes healthy — which
# on a cold boot failed with "this command requires the 'rabbit' app to be
# running on the target node". check_running waits for the application.
test: ["CMD", "rabbitmq-diagnostics", "-q", "check_running"]
interval: 5s
timeout: 10s
retries: 20
# The per-shard `gateway-mock-it-{i}` and `payment-api-it-{i}` services live in
# integration/.it-shards.yaml. The base file's `freight-api-e2e` is never
# started — the app runs in-process (integration/src/app.ts).