Workshop :: Microservices develop, test and deploy

• API gateway
• Distributed Tracing with OpenTelemetry

1. API Gateway with Kong

Kong with Database mode

$git clone https://github.com/Kong/docker-kong
$cd docker-kong/compose/
$KONG_DATABASE=postgres docker compose --profile database up -d

Status of Kong server and database

$docker compose ps
NAME             IMAGE          COMMAND                  SERVICE   CREATED          STATUS                    PORTS
compose-db-1     postgres:9.5   "docker-entrypoint.s…"   db        49 seconds ago   Up 49 seconds (healthy)   5432/tcp
compose-kong-1   kong:latest    "/docker-entrypoint.…"   kong      49 seconds ago   Up 44 seconds (healthy)   0.0.0.0:8000-8002->8000-8002/tcp, [::]:8001-8002->8001-8002/tcp, 0.0.0.0:8443-8444->8443-8444/tcp, [::]:8444->8444/tcp

Gateway API's urls :

• http://localhost:8000 - send traffic to your service via Kong
• http://localhost:8001 - configure Kong using Admin API or via decK
• http://localhost:8002 - access Kong's management Web UI (Kong Manager)

2. Build custom plugin with Kong

$cd kong-plugins
$docker image build -t mykong .

Start Kong server with image=mykong

$docker compose --profile database down
$docker volume prune

$export KONG_DOCKER_TAG=mykong
$KONG_DATABASE=postgres docker compose --profile database up -d
$docker compose ps

List of plugins in Kong manager

• http://localhost:8002/plugins/select

3. Working with Kong API Gateway

• Add service
• Add plugin demo-auth to service
  • http/https
  • URl=http://auth-service:1323/auth
• Add route to service

Add service

$curl http://127.0.0.1:8001/services \
    -d name=demo-api \
    -d url=https://jsonplaceholder.typicode.com

Add route to service

$curl http://127.0.0.1:8001/services/demo-api/routes \
    -d name=demo-api \
	-d 'paths[]=/test'

Check

• https://jsonplaceholder.typicode.com/users/1
• http://localhost:8000/test/users/1

4. Start auth-service

$docker compose build auth-service
$docker compose up -d auth-service

List of urls

• http://localhost:1323/
• http://localhost:1323/auth
• http://localhost:1323/metrics

Check from API Gateway

• http://localhost:8000/test/users/1

5. Add observability for services

• List of services
  • API gateway with Kong
  • Auth service

Add plugins in Kong with Global scope

• http://localhost:8002/plugins
  • New plugin
    • Metric with prometheus
      • http://localhost:8001/metrics
    • Trace and Log with opentelemetry
      • traces_endpoint: http://localhost:4318/v1/traces
      • logs_endpoint: http://localhost:4318/v1/logs

6. Start LGTM Stack

• Log
• Grafana
• Trace
• Metric

$docker compose up -d lgtm
$docker compose up ps

Edit traces_endpoint

• http://lgtm:4318/v1/traces

Check data in Grafana dashboard

• http://localhost:3000

7. Tracing with Python service

• FastAPI
• OpenTelemetry

7.1 Build and run

$docker compose build python-service
$docker compose up -d python-service

List of urls

• http://localhost:9001
• http://localhost:9001/metrics
• http://localhost:9001/health
• http://localhost:9001/articles

7.2 Add python-service to API gateway

• Add service
• Add route to service

Add service

$curl http://127.0.0.1:8001/services \
    -d name=python-service \
    -d url=http://python-service:8000

Add route to service

$curl http://127.0.0.1:8001/services/python-service/routes \
    -d name=python-service \
	-d 'paths[]=/python'

Check from API Gateway

• http://localhost:8000/python/
• http://localhost:8000/python/health

8. Tracing with NodeJS service

• Express
• OpenTelemetry
• Database with PostgreSQL

8.1 Build and run

$docker compose build nodejs-service
$docker compose up -d nodejs-service

List of urls

• http://localhost:9002
• http://localhost:9002/metrics
• http://localhost:9002/call-db
• http://localhost:9002/steps

8.2 Add nodejs-service to API gateway

• Add service
• Add route to service

Add service

$curl http://127.0.0.1:8001/services \
    -d name=nodejs-service \
    -d url=http://nodejs-service:3000

Add route to service

$curl http://127.0.0.1:8001/services/nodejs-service/routes \
    -d name=nodejs-service \
	-d 'paths[]=/nodejs'

Check from API Gateway

• http://localhost:8000/nodejs
• http://localhost:8000/nodejs/health
• http://localhost:8000/nodejs/call-db
• http://localhost:8000/nodejs/steps

9. Application metrics with Prometheus

• API gateway with Kong
  • http://localhost:8001/metrics
• Python service
  • http://localhost:9001/metrics
• NodeJS service
  • http://localhost:9002/metrics

9.1 Install and config Prometheus server

• Config in file prometheus/prometheus.yml
• Start server

File prometheus.yml

scrape_configs:
  - job_name: 'api-gateway'
    scrape_interval: 5s
    metrics_path: '/metrics'
    static_configs:
      - targets: ['kong:8001']
        labels:
          application: 'kong'

  - job_name: 'python-service'
    scrape_interval: 5s
    metrics_path: '/metrics'
    static_configs:
      - targets: ['python-service:8000']
        labels:
          application: 'python-service'
          
  - job_name: 'nodejs-service'
    scrape_interval: 5s
    metrics_path: '/metrics'
    static_configs:
      - targets: ['nodejs-service:3000']
        labels:
          application: 'nodejs-service'

Start server

$docker compose up -d prometheus
$docker compose ps

Access to prometheus

• http://localhost:9090
  • Go to menu Status => Target health