forked from mtusi2026/mtusi2026lab6
Перенесена реализация lab5 (API x2 за HAProxy, PostgreSQL, SeaweedFS S3, RabbitMQ, relay outbox, worker x2, janitor) без изменения бизнес-API. Наблюдаемость: - OpenTelemetry SDK в API, relay, worker, janitor: traces, metrics, logs по OTLP/HTTP; структурные JSON-логи с trace_id/span_id - OTel Collector (memory_limiter, batch, ограниченные очереди и retry), Prometheus (метрики, alert rules), Loki (логи), Tempo (трассы), Grafana с provisioning datasources и dashboard - W3C trace context API -> outbox -> RabbitMQ -> worker (parent-child, включая retry); метрики очереди, пула БД, загрузки worker, CPU/RAM - SLI/SLO, 10 alert rules с unit-тестами promtool и runbooks - инциденты: остановка worker, блокировка БД, отказ telemetry backend; по наблюдениям исправлен простой worker в цикле AMQP (разбор x2.4)
557 lines
23 KiB
YAML
557 lines
23 KiB
YAML
# Lab6 = lab5 + наблюдаемость.
|
||
# HAProxy -> API x N (по умолчанию 2) -> PostgreSQL (данные, сессии, вложения, jobs, outbox)
|
||
# -> SeaweedFS S3 (staging/ и attachments/, приватный bucket)
|
||
# relay: outbox (PostgreSQL) -> RabbitMQ (quorum queue, publisher confirms) + reaper аренд
|
||
# worker x N: RabbitMQ -> staging -> transform (thumbnail/crop/optimized) -> S3 + PostgreSQL
|
||
# telemetry: API/relay/worker/janitor --OTLP/HTTP--> otel-collector --> Prometheus (metrics)
|
||
# --> Loki (logs)
|
||
# --> Tempo (traces)
|
||
# Prometheus также снимает метрики Collector, HAProxy, RabbitMQ, Loki, Tempo;
|
||
# Grafana (127.0.0.1:3000) — provisioning datasources/dashboard из infra/grafana.
|
||
# Запуск: ./scripts/init-secrets.sh && docker compose --env-file .env up -d --build --wait
|
||
name: mtusi-lab6
|
||
|
||
# Один образ на восемь сервисов. Собирает его ТОЛЬКО migrate (первый job стенда), остальные
|
||
# используют готовый тег: параллельная сборка одного тега несколькими сервисами давала
|
||
# «image ... already exists» на пустом Docker.
|
||
x-app-image: &app-image
|
||
image: mtusi-messenger:lab6
|
||
pull_policy: never
|
||
|
||
# Lab6: OpenTelemetry SDK приложений. Экспорт асинхронный, очереди ограничены: недоступный
|
||
# Collector не ломает запросы (теряется только телеметрия). OTEL_SDK_DISABLED=true — выключить.
|
||
x-otel-env: &otel-env
|
||
OTEL_SDK_DISABLED: ${OTEL_SDK_DISABLED:-false}
|
||
OTEL_EXPORTER_OTLP_ENDPOINT: ${OTEL_EXPORTER_OTLP_ENDPOINT:-http://otel-collector:4318}
|
||
OTEL_EXPORTER_OTLP_PROTOCOL: ${OTEL_EXPORTER_OTLP_PROTOCOL:-http/protobuf}
|
||
# Не интерполируется из окружения хоста: Docker CLI сам передаёт compose-плагину
|
||
# OTEL_RESOURCE_ATTRIBUTES=docker.cli.cobra.command_path=... (он попадал в target_info).
|
||
OTEL_RESOURCE_ATTRIBUTES: deployment.environment.name=${APP_ENV:-lab}
|
||
OTEL_TRACES_SAMPLER: ${OTEL_TRACES_SAMPLER:-parentbased_always_on}
|
||
OTEL_TRACES_SAMPLER_ARG: ${OTEL_TRACES_SAMPLER_ARG:-1.0}
|
||
OTEL_METRIC_EXPORT_INTERVAL: ${OTEL_METRIC_EXPORT_INTERVAL:-10000}
|
||
OTEL_EXPORTER_OTLP_TIMEOUT_SECONDS: ${OTEL_EXPORTER_OTLP_TIMEOUT_SECONDS:-5}
|
||
OTEL_BSP_MAX_QUEUE_SIZE: ${OTEL_BSP_MAX_QUEUE_SIZE:-2048}
|
||
OTEL_BLRP_MAX_QUEUE_SIZE: ${OTEL_BLRP_MAX_QUEUE_SIZE:-2048}
|
||
# Собственные метрики SDK: otel.sdk.processor.*.processed{error.type=queue_full} (потери в
|
||
# процессе), otel.sdk.exporter.*.exported{error.type} (ошибки экспорта), размер очереди.
|
||
OTEL_PYTHON_SDK_INTERNAL_METRICS_ENABLED: ${OTEL_PYTHON_SDK_INTERNAL_METRICS_ENABLED:-true}
|
||
|
||
x-db-env: &db-env
|
||
POSTGRES_HOST: db
|
||
POSTGRES_PORT: "5432"
|
||
POSTGRES_DB: ${POSTGRES_DB:-messenger}
|
||
POSTGRES_USER: ${POSTGRES_USER:-messenger}
|
||
POSTGRES_PASSWORD_FILE: /run/secrets/postgres_password
|
||
|
||
x-s3-env: &s3-env
|
||
S3_ENDPOINT: ${S3_ENDPOINT:-http://s3:8333}
|
||
S3_REGION: ${S3_REGION:-us-east-1}
|
||
S3_BUCKET: ${S3_BUCKET:-messenger-attachments}
|
||
S3_CONNECT_TIMEOUT_SECONDS: ${S3_CONNECT_TIMEOUT_SECONDS:-2}
|
||
S3_READ_TIMEOUT_SECONDS: ${S3_READ_TIMEOUT_SECONDS:-5}
|
||
|
||
x-job-env: &job-env
|
||
PIPELINE_VERSION: ${PIPELINE_VERSION:-1}
|
||
JOB_MAX_ATTEMPTS: ${JOB_MAX_ATTEMPTS:-5}
|
||
JOB_TIMEOUT_SECONDS: ${JOB_TIMEOUT_SECONDS:-60}
|
||
JOB_LEASE_SECONDS: ${JOB_LEASE_SECONDS:-30}
|
||
JOB_RETRY_BASE_SECONDS: ${JOB_RETRY_BASE_SECONDS:-2}
|
||
JOB_RETRY_MAX_SECONDS: ${JOB_RETRY_MAX_SECONDS:-60}
|
||
MAX_UPLOAD_BYTES: ${MAX_UPLOAD_BYTES:-10485760}
|
||
MAX_IMAGE_PIXELS: ${MAX_IMAGE_PIXELS:-40000000}
|
||
|
||
services:
|
||
db:
|
||
image: postgres:16.4-alpine@sha256:5660c2cbfea50c7a9127d17dc4e48543eedd3d7a41a595a2dfa572471e37e64c
|
||
environment:
|
||
POSTGRES_DB: ${POSTGRES_DB:-messenger}
|
||
POSTGRES_USER: ${POSTGRES_USER:-messenger}
|
||
POSTGRES_PASSWORD_FILE: /run/secrets/postgres_password
|
||
secrets: [postgres_password]
|
||
volumes:
|
||
- pgdata:/var/lib/postgresql/data
|
||
networks: [backend]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
test: ["CMD-SHELL", "pg_isready -U \"$${POSTGRES_USER}\" -d \"$${POSTGRES_DB}\""]
|
||
interval: 2s
|
||
timeout: 3s
|
||
retries: 30
|
||
stop_grace_period: 30s
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "1.0", memory: 512M}
|
||
|
||
# S3 (как в lab4): master/volume/filer только на 127.0.0.1 контейнера, наружу — S3 API :8333.
|
||
s3:
|
||
image: chrislusf/seaweedfs:4.50@sha256:c2658820e763294a9e9de7b36c81e59b879e3459148351aee4e31a126ca64860
|
||
command:
|
||
- server
|
||
- -dir=/data
|
||
- -ip=127.0.0.1
|
||
- -ip.bind=127.0.0.1
|
||
- -master.volumeSizeLimitMB=1024
|
||
- -master.electionTimeout=2s
|
||
- -volume.preStopSeconds=2
|
||
- -volume.max=0
|
||
- -s3
|
||
- -s3.ip.bind=0.0.0.0
|
||
- -s3.port=8333
|
||
- -s3.config=/run/secrets/s3_identities.json
|
||
- -s3.iam=false
|
||
- -s3.port.iceberg=0
|
||
- -s3.port.lance=0
|
||
- -s3.concurrentUploadLimitMB=64
|
||
- -filer.concurrentUploadLimitMB=64
|
||
- -volume.concurrentUploadLimitMB=64
|
||
- -s3.cacheCapacityMB=16
|
||
- -s3.readerCacheSizeMB=32
|
||
secrets: [s3_identities.json]
|
||
environment:
|
||
GOMEMLIMIT: 512MiB
|
||
volumes:
|
||
- s3data:/data
|
||
networks: [backend]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
test: ["CMD-SHELL", "wget -q -O /dev/null http://127.0.0.1:8333/healthz && curl -sf -m 2 -X PUT --data-binary ok http://127.0.0.1:8888/.healthcheck/probe -o /dev/null"]
|
||
interval: 5s
|
||
start_interval: 1s
|
||
timeout: 3s
|
||
retries: 3
|
||
start_period: 60s
|
||
stop_grace_period: 30s
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "1.0", memory: 768M}
|
||
|
||
# Брокер: RabbitMQ 4.1, quorum-очередь (durable, Raft-журнал на диске), publisher confirms.
|
||
# Топология и пользователи — из definitions (secret). Порты на хост не публикуются.
|
||
broker:
|
||
image: rabbitmq:4.1.4-alpine@sha256:b736d649308e1b3e1a116c3f36986b605ee3d03e88f10166be2900083d2e63f2
|
||
hostname: broker # имя узла rabbit@broker неизменно -> данные volume находятся после пересоздания
|
||
volumes:
|
||
- brokerdata:/var/lib/rabbitmq
|
||
- ./infra/rabbitmq/rabbitmq.conf:/etc/rabbitmq/conf.d/90-lab5.conf:ro
|
||
secrets:
|
||
- source: rabbitmq_definitions.json
|
||
target: /run/secrets/rabbitmq_definitions.json
|
||
# telemetry: Prometheus снимает rabbitmq_prometheus (:15692) — глубина очередей брокера.
|
||
networks: [backend, telemetry]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
# Узел запущен, нет алармов памяти/диска, и рабочая очередь объявлена.
|
||
test: ["CMD-SHELL", "rabbitmq-diagnostics -q check_running && rabbitmq-diagnostics -q check_local_alarms && rabbitmqctl -q -p messenger list_queues name | grep -q attachments.process.v1"]
|
||
interval: 10s
|
||
start_interval: 2s
|
||
timeout: 15s
|
||
retries: 5
|
||
start_period: 60s
|
||
stop_grace_period: 30s
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "1.0", memory: 512M}
|
||
|
||
migrate:
|
||
<<: *app-image
|
||
build:
|
||
context: .
|
||
args:
|
||
APP_VERSION: ${APP_VERSION:-1}
|
||
# Собрать, если образа нет; пересборка — явно (`build` или `up --build`).
|
||
pull_policy: missing
|
||
command: ["python", "/app/src/migrate.py"]
|
||
environment:
|
||
<<: *db-env
|
||
secrets: [postgres_password]
|
||
networks: [backend]
|
||
healthcheck: {disable: true}
|
||
restart: "no"
|
||
depends_on:
|
||
db: {condition: service_healthy}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 128M}
|
||
|
||
s3-init:
|
||
<<: *app-image
|
||
command: ["python", "/app/src/s3_bootstrap.py"]
|
||
environment:
|
||
<<: *s3-env
|
||
S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_admin_access_key
|
||
S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_admin_secret_key
|
||
secrets: [s3_admin_access_key, s3_admin_secret_key]
|
||
networks: [backend]
|
||
healthcheck: {disable: true}
|
||
restart: "no"
|
||
depends_on:
|
||
s3: {condition: service_healthy}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 128M}
|
||
|
||
api:
|
||
<<: *app-image
|
||
expose: ["8080"]
|
||
environment:
|
||
<<: [*db-env, *s3-env, *job-env, *otel-env]
|
||
APP_HOST: 0.0.0.0
|
||
APP_PORT: "8080"
|
||
APP_ENV: ${APP_ENV:-lab}
|
||
SESSION_TTL_SECONDS: ${SESSION_TTL_SECONDS:-3600}
|
||
IDEMPOTENCY_TTL_SECONDS: ${IDEMPOTENCY_TTL_SECONDS:-86400}
|
||
DB_POOL_SIZE: ${DB_POOL_SIZE:-10}
|
||
DRAIN_SECONDS: ${DRAIN_SECONDS:-3}
|
||
SHUTDOWN_TIMEOUT_SECONDS: ${SHUTDOWN_TIMEOUT_SECONDS:-10}
|
||
PASSWORD_HASH_CONCURRENCY: ${PASSWORD_HASH_CONCURRENCY:-2}
|
||
PASSWORD_HASH_WAIT_SECONDS: ${PASSWORD_HASH_WAIT_SECONDS:-0.5}
|
||
DB_HEALTH_INTERVAL_SECONDS: ${DB_HEALTH_INTERVAL_SECONDS:-1}
|
||
READY_STALE_SECONDS: ${READY_STALE_SECONDS:-3}
|
||
S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_api_access_key
|
||
S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_api_secret_key
|
||
UPLOAD_TIMEOUT_SECONDS: ${UPLOAD_TIMEOUT_SECONDS:-60}
|
||
JSON_BODY_TIMEOUT_SECONDS: ${JSON_BODY_TIMEOUT_SECONDS:-10}
|
||
UPLOAD_CONCURRENCY: ${UPLOAD_CONCURRENCY:-8}
|
||
# Проверка JPEG — в подпроцессе src/jpeg_probe.py (аудит F02): progressive JPEG 40 Мпикс
|
||
# 4:4:4 требует до ~250 MiB. Одна проба за раз + её лимит адресного пространства.
|
||
DECODE_CONCURRENCY: ${DECODE_CONCURRENCY:-1}
|
||
JPEG_PROBE_MEMORY_MB: ${JPEG_PROBE_MEMORY_MB:-384}
|
||
JPEG_PROBE_CPU_SECONDS: ${JPEG_PROBE_CPU_SECONDS:-10}
|
||
JPEG_PROBE_TIMEOUT_SECONDS: ${JPEG_PROBE_TIMEOUT_SECONDS:-15}
|
||
S3_PUT_CONCURRENCY: ${S3_PUT_CONCURRENCY:-3}
|
||
FAULT_INJECTION: ${FAULT_INJECTION:-0}
|
||
FAULT_SLOW_HANDLER_SECONDS: ${FAULT_SLOW_HANDLER_SECONDS:-1.5}
|
||
OTEL_SERVICE_NAME: messenger-api
|
||
# У API НЕТ доступа к брокеру: задание пишется в outbox PostgreSQL.
|
||
secrets: [postgres_password, s3_api_access_key, s3_api_secret_key]
|
||
networks: [backend, telemetry]
|
||
restart: unless-stopped
|
||
stop_grace_period: 20s
|
||
depends_on:
|
||
db: {condition: service_healthy}
|
||
migrate: {condition: service_completed_successfully}
|
||
s3-init: {condition: service_completed_successfully}
|
||
deploy:
|
||
replicas: ${API_REPLICAS:-2}
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 512M} # ~100 MiB upload+S3 PUT + проба JPEG до ~260 MiB + запас
|
||
|
||
# Outbox relay + reaper аренд (src/relay.py). Можно масштабировать: SKIP LOCKED + advisory lock.
|
||
relay:
|
||
<<: *app-image
|
||
command: ["python", "/app/src/relay.py"]
|
||
environment:
|
||
<<: [*db-env, *job-env, *otel-env]
|
||
SERVICE_NAME: relay
|
||
BROKER_URL_FILE: /run/secrets/broker_relay_url
|
||
RELAY_POLL_SECONDS: ${RELAY_POLL_SECONDS:-1}
|
||
REAPER_INTERVAL_SECONDS: ${REAPER_INTERVAL_SECONDS:-5}
|
||
JOB_REDRIVE_AFTER_SECONDS: ${JOB_REDRIVE_AFTER_SECONDS:-900}
|
||
OUTBOX_RETENTION_SECONDS: ${OUTBOX_RETENTION_SECONDS:-86400}
|
||
RELAY_FAULT: ${RELAY_FAULT:-}
|
||
OTEL_SERVICE_NAME: messenger-relay
|
||
secrets: [postgres_password, broker_relay_url]
|
||
networks: [backend, telemetry]
|
||
healthcheck:
|
||
test: ["CMD", "python", "-c", "import os,sys,time; sys.exit(0 if time.time() - os.path.getmtime('/tmp/relay.heartbeat') < 30 else 1)"]
|
||
interval: 10s
|
||
timeout: 3s
|
||
retries: 3
|
||
start_period: 30s
|
||
restart: unless-stopped
|
||
stop_grace_period: 15s
|
||
depends_on:
|
||
migrate: {condition: service_completed_successfully}
|
||
broker: {condition: service_healthy}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.25", memory: 128M}
|
||
|
||
# Worker (src/worker.py): WORKER_CONCURRENCY потоков x prefetch 1 на контейнер.
|
||
worker:
|
||
<<: *app-image
|
||
command: ["python", "/app/src/worker.py"]
|
||
environment:
|
||
<<: [*db-env, *s3-env, *job-env, *otel-env]
|
||
SERVICE_NAME: worker
|
||
BROKER_URL_FILE: /run/secrets/broker_worker_url
|
||
S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_worker_access_key
|
||
S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_worker_secret_key
|
||
# Чтение 10 MiB из staging и запись результатов: дольше, чем upload-таймауты API.
|
||
S3_READ_TIMEOUT_SECONDS: ${WORKER_S3_READ_TIMEOUT_SECONDS:-10}
|
||
WORKER_CONCURRENCY: ${WORKER_CONCURRENCY:-2}
|
||
WORKER_PREFETCH: ${WORKER_PREFETCH:-1}
|
||
TRANSFORM_MEMORY_LIMIT_MB: ${TRANSFORM_MEMORY_LIMIT_MB:-768}
|
||
TRANSFORM_CPU_SECONDS: ${TRANSFORM_CPU_SECONDS:-30}
|
||
WORKER_TMP_DIR: /tmp/work
|
||
WORKER_FAULT_INJECTION: ${WORKER_FAULT_INJECTION:-0}
|
||
FAULT_SLOW_SECONDS: ${FAULT_SLOW_SECONDS:-20}
|
||
OTEL_SERVICE_NAME: messenger-worker
|
||
secrets: [postgres_password, broker_worker_url, s3_worker_access_key, s3_worker_secret_key]
|
||
networks: [backend, telemetry]
|
||
healthcheck:
|
||
test: ["CMD", "python", "-c", "import os,sys,time; sys.exit(0 if time.time() - os.path.getmtime('/tmp/worker.heartbeat') < 30 else 1)"]
|
||
interval: 10s
|
||
timeout: 3s
|
||
retries: 3
|
||
start_period: 30s
|
||
restart: unless-stopped
|
||
# SIGTERM: потребители дорабатывают текущее задание (<= JOB_TIMEOUT 60 с) и выходят.
|
||
stop_grace_period: 75s
|
||
depends_on:
|
||
migrate: {condition: service_completed_successfully}
|
||
s3-init: {condition: service_completed_successfully}
|
||
broker: {condition: service_healthy}
|
||
deploy:
|
||
replicas: ${WORKER_REPLICAS:-2}
|
||
resources:
|
||
limits: {cpus: "1.0", memory: 1G}
|
||
|
||
janitor:
|
||
<<: *app-image
|
||
command: ["python", "/app/src/janitor.py"]
|
||
environment:
|
||
<<: [*db-env, *s3-env, *otel-env]
|
||
S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_janitor_access_key
|
||
S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_janitor_secret_key
|
||
ATTACHMENT_GRACE_SECONDS: ${ATTACHMENT_GRACE_SECONDS:-3600}
|
||
JANITOR_INTERVAL_SECONDS: ${JANITOR_INTERVAL_SECONDS:-300}
|
||
STAGING_RETENTION_SECONDS: ${STAGING_RETENTION_SECONDS:-86400}
|
||
OTEL_SERVICE_NAME: messenger-janitor
|
||
secrets: [postgres_password, s3_janitor_access_key, s3_janitor_secret_key]
|
||
networks: [backend, telemetry]
|
||
healthcheck:
|
||
test: ["CMD", "python", "-c", "import os,sys,time; sys.exit(0 if time.time() - os.path.getmtime('/tmp/janitor.heartbeat') < 90 else 1)"]
|
||
interval: 15s
|
||
timeout: 3s
|
||
retries: 2
|
||
start_period: 20s
|
||
restart: unless-stopped
|
||
depends_on:
|
||
migrate: {condition: service_completed_successfully}
|
||
s3-init: {condition: service_completed_successfully}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.25", memory: 128M}
|
||
|
||
lb:
|
||
image: haproxy:3.0.5-alpine@sha256:43a2591a6a62a41dd22edb4452880e9c70de489fbd7c54ffb52d035cccbfe949
|
||
volumes:
|
||
- ./infra/haproxy/haproxy.cfg:/usr/local/etc/haproxy/haproxy.cfg:ro
|
||
environment:
|
||
LOGIN_RATE_LIMIT: ${LOGIN_RATE_LIMIT:-20}
|
||
ports:
|
||
- "127.0.0.1:${LB_PORT:-8080}:8080"
|
||
- "127.0.0.1:8404:8404"
|
||
# telemetry: Prometheus снимает встроенный экспортёр HAProxy (:8404/metrics).
|
||
networks: [frontend, backend, telemetry]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
test: ["CMD", "haproxy", "-c", "-q", "-f", "/usr/local/etc/haproxy/haproxy.cfg"]
|
||
interval: 10s
|
||
timeout: 3s
|
||
retries: 3
|
||
depends_on:
|
||
api: {condition: service_started}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 128M}
|
||
|
||
# Операторский CLI (не часть стенда): docker compose --profile tools run --rm ops <команда>
|
||
# inspect / failed / replay / dlq. Ключ S3 janitor (List, без чтения), брокер — пользователь ops.
|
||
ops:
|
||
<<: *app-image
|
||
profiles: [tools]
|
||
entrypoint: ["python", "/app/src/ops.py"]
|
||
environment:
|
||
<<: [*db-env, *s3-env, *job-env]
|
||
SERVICE_NAME: ops
|
||
BROKER_URL_FILE: /run/secrets/broker_ops_url
|
||
S3_ACCESS_KEY_ID_FILE: /run/secrets/s3_janitor_access_key
|
||
S3_SECRET_ACCESS_KEY_FILE: /run/secrets/s3_janitor_secret_key
|
||
secrets: [postgres_password, broker_ops_url, s3_janitor_access_key, s3_janitor_secret_key]
|
||
networks: [backend]
|
||
healthcheck: {disable: true}
|
||
restart: "no"
|
||
|
||
# Проверки внутри backend-сети (Pillow для попиксельной проверки вариантов, права S3).
|
||
checks:
|
||
<<: *app-image
|
||
profiles: [tools]
|
||
entrypoint: ["python"]
|
||
environment:
|
||
<<: *s3-env
|
||
volumes:
|
||
- ./student-tests:/app/student-tests:ro
|
||
- ./tests:/app/tests:ro
|
||
- ./contracts:/app/contracts:ro
|
||
- ./lab.json:/app/lab.json:ro
|
||
- ./evidence:/app/evidence
|
||
secrets: [s3_api_access_key, s3_api_secret_key, s3_worker_access_key, s3_worker_secret_key,
|
||
s3_janitor_access_key, s3_janitor_secret_key]
|
||
networks: [backend, telemetry]
|
||
healthcheck: {disable: true}
|
||
restart: "no"
|
||
user: "0"
|
||
|
||
# ===================================================================== lab6: телеметрия
|
||
# Collector: получатель/обработчик/экспортёр, не хранилище (буфер в памяти ограничен).
|
||
# Без healthcheck: образ distroless (нет shell); доступность видна по up{job="otel-collector"}.
|
||
otel-collector:
|
||
image: otel/opentelemetry-collector-contrib:0.163.2@sha256:7f22d2a6748110ffb6a59bc0df448bccbf501cdb667d610ede4c61d3863d5849
|
||
command: ["--config=/etc/otelcol/collector.yaml"]
|
||
volumes:
|
||
- ./infra/otel/collector.yaml:/etc/otelcol/collector.yaml:ro
|
||
environment:
|
||
GOMEMLIMIT: 200MiB
|
||
networks: [telemetry]
|
||
restart: unless-stopped
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 256M}
|
||
|
||
# Метрики: Prometheus TSDB (volume promdata), приём OTLP push + scrape; alert rules.
|
||
prometheus:
|
||
image: prom/prometheus:v3.13.5@sha256:aafa9d694e6a883c02db516d9f6e5b84e14195f49a2a6456a4cca815edda72b9
|
||
command:
|
||
- --config.file=/etc/prometheus/prometheus.yml
|
||
- --storage.tsdb.path=/prometheus
|
||
- --storage.tsdb.retention.time=15d
|
||
- --storage.tsdb.retention.size=2GB
|
||
- --web.enable-otlp-receiver
|
||
- --enable-feature=exemplar-storage
|
||
- --web.listen-address=:9090
|
||
volumes:
|
||
- ./infra/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml:ro
|
||
- ./infra/prometheus/rules:/etc/prometheus/rules:ro
|
||
- promdata:/prometheus
|
||
networks: [telemetry]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
test: ["CMD", "wget", "-q", "-O", "/dev/null", "http://127.0.0.1:9090/-/ready"]
|
||
interval: 10s
|
||
start_interval: 2s
|
||
timeout: 3s
|
||
retries: 5
|
||
start_period: 30s
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 512M}
|
||
|
||
# Логи: Loki (volume lokidata), нативный OTLP. Образ distroless (нет shell) — без Docker
|
||
# healthcheck; готовность: up{job="loki"} в Prometheus и scripts/wait-telemetry.sh (/ready).
|
||
loki:
|
||
image: grafana/loki:3.6.19@sha256:0f9cbc340cbe33ee498e654075034336962057db072f57b37f253e14d476443a
|
||
command: ["-config.file=/etc/loki/loki.yaml"]
|
||
volumes:
|
||
- ./infra/loki/loki.yaml:/etc/loki/loki.yaml:ro
|
||
- lokidata:/loki
|
||
networks: [telemetry]
|
||
restart: unless-stopped
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 512M}
|
||
|
||
# Трассы: Tempo (volume tempodata). Distroless, как Loki: готовность — up{job="tempo"}, /ready.
|
||
tempo:
|
||
image: grafana/tempo:2.10.10@sha256:12b0820f13ffa10ce956b6e9acc3b03f3cffded99109be6cc09d106fe56e5d38
|
||
command: ["-config.file=/etc/tempo/tempo.yaml"]
|
||
volumes:
|
||
- ./infra/tempo/tempo.yaml:/etc/tempo/tempo.yaml:ro
|
||
- tempodata:/var/tempo
|
||
networks: [telemetry]
|
||
restart: unless-stopped
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 512M}
|
||
|
||
# Интерфейс: Grafana (только 127.0.0.1:3000). Пароль admin — Docker secret.
|
||
grafana:
|
||
image: grafana/grafana:12.4.14@sha256:ae509d9ae8c1021ca32ec370a9ca975d40165eda9fccd1e1102d317ff8787cd7
|
||
environment:
|
||
GF_SECURITY_ADMIN_USER: admin
|
||
GF_SECURITY_ADMIN_PASSWORD__FILE: /run/secrets/grafana_admin_password
|
||
GF_USERS_ALLOW_SIGN_UP: "false"
|
||
GF_AUTH_ANONYMOUS_ENABLED: "false"
|
||
GF_ANALYTICS_REPORTING_ENABLED: "false"
|
||
GF_ANALYTICS_CHECK_FOR_UPDATES: "false"
|
||
GF_ANALYTICS_CHECK_FOR_PLUGIN_UPDATES: "false"
|
||
GF_NEWS_NEWS_FEED_ENABLED: "false"
|
||
GF_PLUGINS_PREINSTALL_DISABLED: "true"
|
||
GF_DASHBOARDS_DEFAULT_HOME_DASHBOARD_PATH: /var/lib/grafana/dashboards/messenger.json
|
||
secrets: [grafana_admin_password]
|
||
volumes:
|
||
- ./infra/grafana/provisioning:/etc/grafana/provisioning:ro
|
||
- ./infra/grafana/dashboards:/var/lib/grafana/dashboards:ro
|
||
- grafanadata:/var/lib/grafana
|
||
ports:
|
||
- "127.0.0.1:${GRAFANA_PORT:-3000}:3000"
|
||
networks: [telemetry, grafana-ui]
|
||
restart: unless-stopped
|
||
healthcheck:
|
||
test: ["CMD", "wget", "-q", "-O", "/dev/null", "http://127.0.0.1:3000/api/health"]
|
||
interval: 10s
|
||
start_interval: 2s
|
||
timeout: 3s
|
||
retries: 6
|
||
start_period: 60s
|
||
depends_on:
|
||
prometheus: {condition: service_started}
|
||
loki: {condition: service_started}
|
||
tempo: {condition: service_started}
|
||
deploy:
|
||
resources:
|
||
limits: {cpus: "0.5", memory: 384M}
|
||
|
||
secrets:
|
||
grafana_admin_password:
|
||
file: ./secrets/grafana_admin_password
|
||
postgres_password:
|
||
file: ./secrets/postgres_password
|
||
s3_identities.json:
|
||
file: ./secrets/s3_identities.json
|
||
s3_admin_access_key:
|
||
file: ./secrets/s3_admin_access_key
|
||
s3_admin_secret_key:
|
||
file: ./secrets/s3_admin_secret_key
|
||
s3_api_access_key:
|
||
file: ./secrets/s3_api_access_key
|
||
s3_api_secret_key:
|
||
file: ./secrets/s3_api_secret_key
|
||
s3_worker_access_key:
|
||
file: ./secrets/s3_worker_access_key
|
||
s3_worker_secret_key:
|
||
file: ./secrets/s3_worker_secret_key
|
||
s3_janitor_access_key:
|
||
file: ./secrets/s3_janitor_access_key
|
||
s3_janitor_secret_key:
|
||
file: ./secrets/s3_janitor_secret_key
|
||
rabbitmq_definitions.json:
|
||
file: ./secrets/rabbitmq_definitions.json
|
||
broker_relay_url:
|
||
file: ./secrets/broker_relay_url
|
||
broker_worker_url:
|
||
file: ./secrets/broker_worker_url
|
||
broker_ops_url:
|
||
file: ./secrets/broker_ops_url
|
||
|
||
networks:
|
||
frontend: {}
|
||
backend:
|
||
internal: true
|
||
# Lab6: сеть телеметрии изолирована от данных (Grafana не видит БД/S3/секреты приложений).
|
||
telemetry:
|
||
internal: true
|
||
# Только для публикации Grafana на 127.0.0.1.
|
||
grafana-ui: {}
|
||
|
||
volumes:
|
||
pgdata: {}
|
||
s3data: {}
|
||
brokerdata: {}
|
||
promdata: {}
|
||
lokidata: {}
|
||
tempodata: {}
|
||
grafanadata: {}
|