k6 with InfluxDB(& Prometheus)

이 글에서는 k6의 결과물을 외부 서비스로 스트리밍하여 시계열 데이터를 시각화하는 방법을 알아봅니다.

• (1/4) 성능 테스트

• (2/4) Get started with k6

• (4/4) 로그 수집 서비스 성능 개선하기

결론부터 말하자면, InfluxDB의 버전 이슈로 InfluxDB보다는 Prometheus로 설정하는 것이 훨씬 나았습니다. InfluxDB로 설정하는 다양한 방법을 찾아서 시도해보았지만 어느 것도 만족스럽지 못했습니다. 시각화 방법만 궁금하신 분은 Prometheus 부분부터 보시면 됩니다.

InfluxDB

공식 문서 버전

# Install xk6
go install go.k6.io/xk6/cmd/xk6@latest

# Build the k6 binary
xk6 build --with github.com/grafana/xk6-output-influxdb

docker run --rm -it -e GOOS=darwin -u "$(id -u):$(id -g)" -v "${PWD}:/xk6" \
  grafana/xk6 build v0.43.1 \
  --with github.com/mostafa/xk6-kafka@v0.17.0 \
  --with github.com/grafana/xk6-output-influxdb@v0.3.0

brew install influxdb
brew services start influxdb

K6_INFLUXDB_ORGANIZATION="<INFLUXDB-ORGANIZATION-NAME>" \
K6_INFLUXDB_BUCKET="<INFLUXDB-BUCKET-NAME>" \
K6_INFLUXDB_TOKEN="<INFLUXDB-TOKEN>" \
K6_INFLUXDB_ADDR="<INFLUXDB-HTTP-ADDRESS>" \
./k6 run script.js -o xk6-influxdb

brew update
brew install grafana
brew services start grafana

더 간단한 방법(with. Docker-Compose)

./docker-run.sh samples/http_2.js

하지만…

Why is this output not directly part of k6 core? The k6 core already supports the InfluxDB v1 so the natural feeling would be to do the same for the v2. Unfortunately, the v2 has introduced some breaking changes in the core parts of the API. This would make it difficult to support both versions without taking a bunch of compromises for maintaining the retro-compatibility or introducing breaking changes in the current user experience of the k6's InfluxDB output, with a high probability to create more confusion for k6's users. For this main reason, the k6 development team has decided to create a new and independent extension for InfluxDB v2. [xk6-output-influxdb README]

Prometheus

K6_PROMETHEUS_RW_SERVER_URL=http://localhost:9090/api/v1/write \
K6_PROMETHEUS_RW_USERNAME=USERNAME \
K6_PROMETHEUS_RW_PASSWORD=PASSWORD \
k6 run -o experimental-prometheus-rw script.js

docker-compose

// docker-compose.yml
services:
  grafana:
    image: grafana/grafana
    ports:
      - 3000:3000
    volumes:
      - ./grafana/provisioning:/etc/grafana/provisioning
      - ./grafana/dashboard:/dashboard
    depends_on:
      - prometheus
  prometheus:
    image: prom/prometheus
    ports:
      - 9090:9090
    command:
      - --web.enable-remote-write-receiver
      - --enable-feature=native-histograms
      - --config.file=/etc/prometheus/prometheus.yml
  k6:
    image: grafana/k6
    volumes:
      - ./script:/scripts
    command: run -o experimental-prometheus-rw /scripts/script.js
    environment:
      - K6_PROMETHEUS_RW_SERVER_URL=http://prometheus:9090/api/v1/write
      - K6_PROMETHEUS_RW_TREND_AS_NATIVE_HISTOGRAM=true
    depends_on:
      - prometheus

command:
  - --enable-feature=native-histograms

environment:
    - K6_PROMETHEUS_RW_TREND_AS_NATIVE_HISTOGRAM=true

• Counter and Gauge metrics (default)

• Prometheus Native histogram

// prometheus.yml
apiVersion: 1

datasources:
  - name: Prometheus
    type: prometheus
    access: proxy
    url: http://prometheus:9090

// dashboards.yml
apiVersion: 1

providers:
  - name: "k6 Test"
    orgId: 1
    folder: ""
    type: file
    disableDeletion: false
    updateIntervalSeconds: 10
    options:
      path: /dashboard/18030_rev8.json
      foldersFromFilesStructure: true

정리