Files
claudemesh/docker-compose.production.yml
Alejandro Gutiérrez a4cd068ef5
Some checks failed
CI / Tests / 🧪 Test (push) Has been cancelled
feat(deploy): pre-start drizzle-kit migrate init container
One-shot migrate container runs drizzle-kit migrate against DATABASE_URL
and exits 0 before web boots. web service depends_on with condition
service_completed_successfully, so failed migrations block web startup
instead of serving 500s against a stale schema. Broker deliberately does
NOT depend on migrate - it tolerates DB-down gracefully per DEPLOY_SPEC
and should keep serving WS peers even during migration failures.

Also excludes apps/cli from docker build context (CLI ships to npm, not
containers) to sidestep zod spec drift in its package.json vs lockfile.

Known followup: migrate image is 3.27GB due to pnpm catalog: specifiers
forcing full-workspace resolution. pnpm deploy bundle trim is a P2.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-05 14:55:36 +01:00

95 lines
3.2 KiB
YAML

# claudemesh — production compose (for Coolify Service deployment)
#
# Three services:
# - migrate → one-shot drizzle-kit migrate, exits 0, gates web startup
# - broker → ic.claudemesh.com (WSS /ws + HTTP /health + /hook/set-status)
# - web → claudemesh.com + dashboard.claudemesh.com (Next.js)
#
# Postgres is NOT declared here — managed externally by Coolify or a managed DB.
# Pass DATABASE_URL + all secrets at runtime via Coolify env config.
#
# Why broker does NOT depend on migrate:
# Broker tolerates DB-down gracefully (per apps/broker/DEPLOY_SPEC.md §Healthcheck).
# It should keep serving even if a migration is in-flight or has failed, so WS
# peers stay connected + /health reports degraded instead of going 502.
#
# Why web DOES depend on migrate:
# Next.js routes assume the schema they were built against. Starting web before
# migrations land → 500s on every query touching new tables/columns.
name: claudemesh
services:
migrate:
image: ${MIGRATE_IMAGE:-claudemesh-migrate:latest}
restart: "no"
environment:
DATABASE_URL: ${DATABASE_URL}
networks:
- claudemesh-internal
broker:
image: ${BROKER_IMAGE:-claudemesh-broker:latest}
restart: always
environment:
NODE_ENV: production
BROKER_PORT: 7900
DATABASE_URL: ${DATABASE_URL}
STATUS_TTL_SECONDS: ${STATUS_TTL_SECONDS:-60}
HOOK_FRESH_WINDOW_SECONDS: ${HOOK_FRESH_WINDOW_SECONDS:-30}
MAX_CONNECTIONS_PER_MESH: ${MAX_CONNECTIONS_PER_MESH:-100}
MAX_MESSAGE_BYTES: ${MAX_MESSAGE_BYTES:-65536}
HOOK_RATE_LIMIT_PER_MIN: ${HOOK_RATE_LIMIT_PER_MIN:-30}
expose:
- "7900"
networks:
- coolify
- claudemesh-internal
healthcheck:
test: ["CMD", "bun", "-e", "fetch('http://localhost:7900/health').then(r=>{process.exit(r.ok?0:1)}).catch(()=>process.exit(1))"]
interval: 15s
timeout: 5s
start_period: 10s
retries: 3
web:
image: ${WEB_IMAGE:-claudemesh-web:latest}
restart: always
environment:
NODE_ENV: production
PORT: 3000
HOSTNAME: 0.0.0.0
DATABASE_URL: ${DATABASE_URL}
BETTER_AUTH_SECRET: ${BETTER_AUTH_SECRET}
BETTER_AUTH_URL: ${BETTER_AUTH_URL:-https://claudemesh.com}
BETTER_AUTH_TRUSTED_ORIGINS: ${BETTER_AUTH_TRUSTED_ORIGINS:-https://claudemesh.com,https://dashboard.claudemesh.com,https://ic.claudemesh.com}
GITHUB_CLIENT_ID: ${GITHUB_CLIENT_ID:-}
GITHUB_CLIENT_SECRET: ${GITHUB_CLIENT_SECRET:-}
GOOGLE_CLIENT_ID: ${GOOGLE_CLIENT_ID:-}
GOOGLE_CLIENT_SECRET: ${GOOGLE_CLIENT_SECRET:-}
BROKER_INTERNAL_URL: http://broker:7900
expose:
- "3000"
networks:
- coolify
- claudemesh-internal
depends_on:
migrate:
condition: service_completed_successfully
broker:
condition: service_healthy
healthcheck:
test: ["CMD", "node", "-e", "fetch('http://localhost:3000').then(r=>{process.exit(r.ok?0:1)}).catch(()=>process.exit(1))"]
interval: 15s
timeout: 5s
start_period: 20s
retries: 3
networks:
# Coolify's shared Traefik network — must already exist on the host
coolify:
external: true
# Internal backplane between migrate + broker + web
claudemesh-internal:
driver: bridge