Distributed Locust Example

info

This Workflows functionality is not available when running the Testkube Agent in Standalone Mode - Read More

Testkube has built-in support for parallelising any testing tool - check out the Parallelization documentation for details and examples.

The below example shows how to distribute a Locust test across a configurable number of "workers"

Distributed JMeter Workflow

kind: TestWorkflow
apiVersion: testworkflows.testkube.io/v1
metadata:
  name: distributed-locust-example
  namespace: testkube
  labels:
    docs: example
spec:
  config:
    workers:
      type: integer
      default: 2
  content:
    git:
      uri: https://github.com/kubeshop/testkube
      revision: main
      paths:
      - test/locust/locust-project
  services:
    master:
      count: 1
      logs: always
      timeout: 30s
      transfer:
      - from: /data/repo/test/locust/locust-project
      image: locustio/locust:2.32.3
      shell: |
        mkdir /data/artifacts
        cd /data/repo/test/locust/locust-project
        locust --master -f locust_test.py --host https://testkube-test-page-lipsum.pages.dev/ --headless \
        --users 20 --spawn-rate 10 --run-time 60s --stop-timeout 180 --html=/data/artifacts/report.html
      readinessProbe:
        tcpSocket:
          port: 5557
        periodSeconds: 1
  steps:
  - name: Run test
    parallel:
      count: config.workers
      transfer:
      - from: /data/repo
      use:
      - name: distribute/evenly
      container:
        workingDir: /data/repo/test/locust/locust-project
      paused: true
      run:
        image: locustio/locust:2.32.3
        shell: locust -f - --worker --master-host {{ services.master.0.ip }} --processes
          {{ config.workers }}

After execution, you can see the output from the test executions under the executions panel tabs:

Log Output

The log output from the Locust execution: