# Docker Swarm stack лабораторной 5. # # Роли: Traefik (единственная публичная точка входа), api (2 взаимозаменяемые # реплики), PostgreSQL (метаданные, jobs/outbox/variants), Valkey (сессии), # MinIO (приватные байты и staging), RabbitMQ (durable quorum queue + DLQ), # relay (transactional outbox + reaper lease), worker (обработка изображений), # minio-init (bootstrap bucket/прав), migrate (однократная схема). # Наружу опубликован только Traefik :8080; API, PostgreSQL, Valkey, MinIO и # RabbitMQ портов не публикуют. # # Версии образов закреплены тегом и digest-ом; для приложения тег задаёт # scripts/deploy.sh. Compose-вариант не заявляется: выбран Swarm. x-api-environment: &api-environment APP_ENV: lab APP_HOST: 0.0.0.0 APP_PORT: "8080" LOG_LEVEL: info SESSION_STORE: redis SESSION_TTL_SECONDS: "${SESSION_TTL_SECONDS:-3600}" DB_POOL_SIZE: "5" DB_MAX_OVERFLOW: "5" DB_POOL_TIMEOUT_SECONDS: "2" DB_CONNECT_TIMEOUT_SECONDS: "1.5" DB_COMMAND_TIMEOUT_SECONDS: "2.5" DB_STATEMENT_TIMEOUT_MS: "2000" CACHE_TIMEOUT_SECONDS: "1" GRACEFUL_SHUTDOWN_SECONDS: "5" SHUTDOWN_DRAIN_SECONDS: "2" S3_ENDPOINT: "http://minio:9000" S3_REGION: "us-east-1" S3_BUCKET: "messenger-attachments" S3_KEY_PREFIX: "attachments/" S3_CONNECT_TIMEOUT_SECONDS: "1.5" S3_READ_TIMEOUT_SECONDS: "30" S3_MAX_ATTEMPTS: "2" MAX_UPLOAD_BYTES: "10485760" MAX_IMAGE_PIXELS: "40000000" ATTACHMENT_ORPHAN_GRACE_SECONDS: "3600" STAGING_RETENTION_SECONDS: "86400" ATTACHMENT_FAULT_MODE: "none" PIPELINE_VERSION: "2" JOB_MAX_ATTEMPTS: "5" ALLOW_FAULT_INJECTION: "true" DATABASE_URL_FILE: /run/secrets/database_url CACHE_URL_FILE: /run/secrets/cache_url BROKER_URL_FILE: /run/secrets/broker_url CURSOR_SECRET_FILE: /run/secrets/cursor_secret S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_access_key_id S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_secret_access_key x-worker-environment: &worker-environment APP_ENV: lab LOG_LEVEL: info DB_POOL_SIZE: "3" DB_MAX_OVERFLOW: "3" DB_POOL_TIMEOUT_SECONDS: "2" DB_CONNECT_TIMEOUT_SECONDS: "1.5" DB_COMMAND_TIMEOUT_SECONDS: "5" DB_STATEMENT_TIMEOUT_MS: "10000" CACHE_TIMEOUT_SECONDS: "1" GRACEFUL_SHUTDOWN_SECONDS: "10" S3_ENDPOINT: "http://minio:9000" S3_REGION: "us-east-1" S3_BUCKET: "messenger-attachments" S3_KEY_PREFIX: "attachments/" S3_CONNECT_TIMEOUT_SECONDS: "1.5" S3_READ_TIMEOUT_SECONDS: "60" S3_MAX_ATTEMPTS: "3" MAX_UPLOAD_BYTES: "10485760" MAX_IMAGE_PIXELS: "40000000" PIPELINE_VERSION: "2" WORKER_CONCURRENCY: "${WORKER_CONCURRENCY:-2}" WORKER_PREFETCH: "${WORKER_PREFETCH:-2}" JOB_MAX_ATTEMPTS: "${JOB_MAX_ATTEMPTS:-5}" JOB_TIMEOUT_SECONDS: "${JOB_TIMEOUT_SECONDS:-60}" JOB_LEASE_SECONDS: "${JOB_LEASE_SECONDS:-15}" JOB_BACKOFF_BASE_SECONDS: "1" JOB_BACKOFF_MAX_SECONDS: "30" STAGING_RETENTION_SECONDS: "86400" ALLOW_FAULT_INJECTION: "true" RELAY_POLL_SECONDS: "0.25" REAPER_INTERVAL_SECONDS: "1" DATABASE_URL_FILE: /run/secrets/database_url CACHE_URL_FILE: /run/secrets/cache_url BROKER_URL_FILE: /run/secrets/broker_url S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_access_key_id S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_secret_access_key services: traefik: image: traefik:v3.5.4 command: - --providers.swarm=true - --providers.swarm.endpoint=unix:///var/run/docker.sock - --providers.swarm.exposedByDefault=false - --providers.swarm.network=backend - --providers.swarm.watch=true - --entryPoints.web.address=:8080 - --log.level=INFO - --accesslog=false ports: - target: 8080 published: 8080 protocol: tcp mode: ingress volumes: - /var/run/docker.sock:/var/run/docker.sock:ro networks: - backend deploy: replicas: 1 placement: constraints: - node.role == manager resources: limits: cpus: "0.50" memory: 128M restart_policy: condition: any delay: 5s postgres: image: postgres:16.15-bookworm@sha256:0ea6700a3b4f0ae6ce746519073558aed4d88a79d8d07622a9a644946c7319c4 environment: POSTGRES_USER: messenger POSTGRES_DB: messenger POSTGRES_PASSWORD_FILE: /run/secrets/postgres_password POSTGRES_INITDB_ARGS: "--auth-host=scram-sha-256" secrets: - source: lab5_postgres_password target: postgres_password volumes: - pgdata:/var/lib/postgresql/data networks: - backend healthcheck: test: ["CMD-SHELL", "pg_isready -U messenger -d messenger"] interval: 5s timeout: 3s retries: 12 start_period: 10s deploy: replicas: 1 resources: limits: cpus: "2.0" memory: 512M restart_policy: condition: any delay: 5s valkey: image: valkey/valkey:8.0.11-trixie@sha256:4436c94fc34ce4af0354b9379d433a1f258998fb955f995c89822f98c14a8cab command: - valkey-server - --appendonly - "yes" - --appendfsync - everysec - --save - "" volumes: - valkeydata:/data networks: - backend healthcheck: test: ["CMD", "valkey-cli", "ping"] interval: 5s timeout: 3s retries: 12 start_period: 5s deploy: replicas: 1 resources: limits: cpus: "1.0" memory: 256M restart_policy: condition: any delay: 5s rabbitmq: # Версия фиксируется digest-ом (см. evidence/04-infrastructure.txt). image: rabbitmq:4.1.4-alpine@sha256:b736d649308e1b3e1a116c3f36986b605ee3d03e88f10166be2900083d2e63f2 entrypoint: ["/bin/sh", "-c"] command: - >- export RABBITMQ_DEFAULT_USER=$$(cat /run/secrets/rabbitmq_user); export RABBITMQ_DEFAULT_PASS=$$(cat /run/secrets/rabbitmq_password); exec docker-entrypoint.sh rabbitmq-server environment: # Стабильный node name: имя контейнера в Swarm меняется, а mnesia-каталог # должен переживать restart вместе с volume. RABBITMQ_NODENAME: "rabbit@localhost" RABBITMQ_USE_LONGNAME: "false" RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS: "+S 2:2" secrets: - source: lab5_rabbitmq_user target: rabbitmq_user - source: lab5_rabbitmq_password target: rabbitmq_password volumes: - rabbitmqdata:/var/lib/rabbitmq networks: - backend healthcheck: test: ["CMD", "rabbitmq-diagnostics", "-q", "ping"] interval: 10s timeout: 5s retries: 18 start_period: 30s deploy: replicas: 1 resources: limits: cpus: "1.0" memory: 512M restart_policy: condition: any delay: 5s minio: # RELEASE.2025-09-07T16-13-09Z; digest фиксирует именно проверенный образ. image: minio/minio@sha256:14cea493d9a34af32f524e538b8346cf79f3321eff8e708c1e2960462bd8936e entrypoint: ["/bin/sh", "-c"] command: - >- export MINIO_ROOT_USER=$$(cat /run/secrets/s3_root_user); export MINIO_ROOT_PASSWORD=$$(cat /run/secrets/s3_root_password); exec /usr/bin/minio server /data --console-address :9001 environment: MINIO_BROWSER: "off" secrets: - source: lab5_s3_root_user target: s3_root_user - source: lab5_s3_root_password target: s3_root_password volumes: - miniodata:/data networks: - backend healthcheck: test: ["CMD", "curl", "-f", "http://127.0.0.1:9000/minio/health/live"] interval: 5s timeout: 3s retries: 12 start_period: 10s deploy: replicas: 1 resources: limits: cpus: "1.0" memory: 512M restart_policy: condition: any delay: 5s minio-init: # Тот же проверенный MinIO-образ: в нём уже есть /bin/sh и mc. image: minio/minio@sha256:14cea493d9a34af32f524e538b8346cf79f3321eff8e708c1e2960462bd8936e entrypoint: ["/bin/sh", "/etc/minio-init.sh"] environment: S3_ENDPOINT: "http://minio:9000" S3_BUCKET: "messenger-attachments" S3_POLICY_NAME: "messenger-attachments-rw" MC_CONFIG_DIR: "/tmp/.mc" MINIO_ROOT_USER_FILE: /run/secrets/s3_root_user MINIO_ROOT_PASSWORD_FILE: /run/secrets/s3_root_password S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_access_key_id S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_secret_access_key secrets: - source: lab5_s3_root_user target: s3_root_user - source: lab5_s3_root_password target: s3_root_password - source: lab5_s3_access_key_id target: s3_access_key_id - source: lab5_s3_secret_access_key target: s3_secret_access_key configs: - source: lab5_minio_init target: /etc/minio-init.sh mode: 0555 networks: - backend deploy: replicas: 1 restart_policy: condition: none resources: limits: cpus: "0.5" memory: 128M migrate: image: ${API_IMAGE:-mtusi-messenger-lab5:5.0.0} command: - /bin/sh - -c - "until alembic upgrade head; do echo migration waiting for postgres; sleep 2; done" environment: DATABASE_URL_FILE: /run/secrets/database_url secrets: - source: lab5_database_url target: database_url networks: - backend deploy: replicas: 1 resources: limits: cpus: "1.0" memory: 256M restart_policy: condition: none api: image: ${API_IMAGE:-mtusi-messenger-lab5:5.0.0} environment: <<: *api-environment secrets: - source: lab5_database_url target: database_url - source: lab5_cache_url target: cache_url - source: lab5_broker_url target: broker_url - source: lab5_cursor_secret target: cursor_secret - source: lab5_s3_access_key_id target: s3_access_key_id - source: lab5_s3_secret_access_key target: s3_secret_access_key networks: - backend stop_grace_period: 15s healthcheck: test: ["CMD", "python", "-m", "messenger.healthcheck"] interval: 5s timeout: 3s start_period: 15s retries: 3 deploy: replicas: 2 labels: - traefik.enable=true - traefik.swarm.network=backend - traefik.http.routers.messenger.rule=PathPrefix(`/`) - traefik.http.routers.messenger.entrypoints=web - traefik.http.routers.messenger.service=messenger - traefik.http.services.messenger.loadbalancer.server.port=8080 - traefik.http.services.messenger.loadbalancer.server.scheme=http - traefik.http.services.messenger.loadbalancer.healthcheck.path=/health/ready - traefik.http.services.messenger.loadbalancer.healthcheck.interval=2s - traefik.http.services.messenger.loadbalancer.healthcheck.unhealthyinterval=1s - traefik.http.services.messenger.loadbalancer.healthcheck.timeout=1s - traefik.http.services.messenger.loadbalancer.healthcheck.scheme=http update_config: parallelism: 1 order: start-first delay: 5s monitor: 10s failure_action: rollback rollback_config: parallelism: 1 order: start-first delay: 5s restart_policy: condition: any delay: 1s max_attempts: 0 resources: limits: cpus: "1.0" memory: 512M reservations: cpus: "0.25" memory: 128M relay: image: ${API_IMAGE:-mtusi-messenger-lab5:5.0.0} command: ["python", "-m", "messenger.relay"] environment: <<: *worker-environment HEARTBEAT_FILE: /tmp/messenger-relay-heartbeat secrets: - source: lab5_database_url target: database_url - source: lab5_cache_url target: cache_url - source: lab5_broker_url target: broker_url - source: lab5_s3_access_key_id target: s3_access_key_id - source: lab5_s3_secret_access_key target: s3_secret_access_key networks: - backend stop_grace_period: 15s healthcheck: test: ["CMD", "python", "-m", "messenger.worker_health"] interval: 10s timeout: 3s start_period: 10s retries: 3 deploy: replicas: 1 restart_policy: condition: any delay: 1s resources: limits: cpus: "0.5" memory: 256M worker: image: ${API_IMAGE:-mtusi-messenger-lab5:5.0.0} command: ["python", "-m", "messenger.worker"] environment: <<: *worker-environment HEARTBEAT_FILE: /tmp/messenger-worker-heartbeat secrets: - source: lab5_database_url target: database_url - source: lab5_cache_url target: cache_url - source: lab5_broker_url target: broker_url - source: lab5_s3_access_key_id target: s3_access_key_id - source: lab5_s3_secret_access_key target: s3_secret_access_key networks: - backend stop_grace_period: 20s healthcheck: test: ["CMD", "python", "-m", "messenger.worker_health"] interval: 10s timeout: 3s start_period: 10s retries: 3 deploy: replicas: 1 restart_policy: condition: any delay: 1s resources: limits: cpus: "1.0" memory: 512M configs: lab5_minio_init: file: ./scripts/minio-init.sh secrets: lab5_database_url: external: true lab5_cache_url: external: true lab5_broker_url: external: true lab5_postgres_password: external: true lab5_cursor_secret: external: true lab5_rabbitmq_user: external: true lab5_rabbitmq_password: external: true lab5_s3_root_user: external: true lab5_s3_root_password: external: true lab5_s3_access_key_id: external: true lab5_s3_secret_access_key: external: true volumes: pgdata: valkeydata: miniodata: rabbitmqdata: networks: backend: driver: overlay attachable: true