Distributed load testing using kubernetes and locust

The repository contains everything needed to run a distributed load testing environment in kubernetes using a locust master and locust slaves.

Table of contents

1. Project structure
2. Prerequisites
3. Running
   1. tl;dr
   2. Detailed installation
      1. Preparing the local kubernetes cluster
      2. Building the docker image
      3. Installing the helm charts
      4. Confirm the installation and access locust dashboard
4. Architecture

Project structure

.
├── .vscode                                 # launch.json and tasks.json needed to debug the tasks.py in vsc
├── docker-image                            # The shared docker image for locust masters and slaves
│   ├── locust-tasks                        # Python source code
|   |   ├── requirements.txt                # Python dependencies for tasks.py
|   |   ├── tasks.py                        # Locust tasks
|   ├── run.sh                              # Shell script to determine if the docker containers should be master or slave
│   ├── Dockerfile                          # Dockerfile
├── loadtest-chart                          # Helm chart
|   ├── templates                           # Helm templates
|   |   ├── _helpers.tpl                    # Helm helpers
|   |   ├── locust-master-deployment.yaml   # Kubernetes deployment configuration for locust master
|   |   ├── locust-master-service.yaml      # Kubernetes service configuration for locust master
|   |   ├── locust-slave-deployment.yaml    # Kubernetes deployment configuration for locust slaves
|   ├── Chart.yaml                          # Chart definition
|   ├── values.yaml                         # Chart definition
└── ...

---

Prerequisites

Product	Version	Link
Python	2.7.15	Windows, MacOS
Docker	18.03.0-ce (23751)	Windows, MacOS
kubectl	2.0.0	Windows, MacOS
Minikube	0.27.0	Windows, MacOS
helm	2.9.1	Windows MacOS

Running

Deploy locus project

helm install directory-locust-kubernetes/loadtest-chart/ --namespace locust --name locust --set hostAliases.ip=${web_ip}

Start swarm by command lines

Locally :

curl -X POST -F 'locust_count=500' -F 'hatch_rate=5' http://<locust-url>:<locust-port>/swarm

In kubernetes :

kubectl run --generator=run-pod/v1 startlocust --image=djbingham/curl --restart='OnFailure' -i --tty --rm --command -- curl -X POST -F 'locust_count=500' -F 'hatch_rate=10' http://locust-master.locust.svc.cluster.local:8089/swarm

Stop load test by command lines

Locally:

curl http://<locust-url>:<locust-port>/stop

In kubernetes :

kubectl run --generator=run-pod/v1 stoplocuts --generator=run-pod/v1 --image=djbingham/curl --restart='OnFailure' -i --tty --rm --command -- curl http://locust-master.locust.svc.cluster.local:8089/stop

---

Detailed installation

Building the docker image

In the root of the repo, run docker build docker-image -t locust-tasks:latest

Architecture