> OctoPerf load testing - How to integrate OctoPerf into Gitlab CI.

Source: https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/ci/azure-devops/

# Azure Devops

![Azure](https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/ci/img/microsoft-azure.png)

**Azure DevOps** is a [Microsoft](https://en.wikipedia.org/wiki/Microsoft) product that provides a variety of services to covers the entire [application lifecycle](https://en.wikipedia.org/wiki/Application_Lifecycle_Management), and enables [DevOps](https://en.wikipedia.org/wiki/DevOps) capabilities.

OctoPerf seamlessly integrates with Azure Devops thanks to our [Maven Plugin](https://github.com/OctoPerf/octoperf-maven-plugin). As part of your software builds, you can run OctoPerf tasks like:

- Run an existing scenario,
- Import and run a [JMeter JMX](https://jmeter.apache.org) stored in your git repository,
- Or even stop the build automatically whenever a [threshold or SLA](https://api.octoperf.com/doc/monitoring/create-connection/sla/) exceeds a certain criticality (Example: `WARNING`).

## Setup

### Prerequisites

In order to setup Azure devops builds to run OctoPerf load tests, first you need:

- An [Azure](https://azure.microsoft.com/) account,
- And an [OctoPerf](https://api.octoperf.com/) account.

### Azure pipeline

The first step is to get a pipeline ready in Azure Devops. A complete guide can be found [in their documentation](https://docs.microsoft.com/fr-fr/azure/devops/pipelines/ecosystems/java?view=azure-devops).

Basically you need to create a pipeline and construct an `azure-pipelines.yml` file, in our case we will use something like this:

```yml
steps:
- task: Maven@3
  inputs:
    mavenPomFile: 'pom.xml'
    mavenOptions: '-Xmx512m'
    javaHomeOption: 'JDKVersion'
    jdkVersionOption: '1.11'
    jdkArchitectureOption: 'x64'
    publishJUnitResults: false
    goals: 'octoperf:execute-scenario'
```

### OctoPerf Maven plugin

You can find the OctoPerf Maven plugin configuration [on its documentation page](https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/maven/#configuration).

To summarize, you need to provide a `pom.xml` file and adapt the following section:

```xml
<apiKey>YOUR_API_KEY</apiKey>
<workspaceName>WORKSPACE_NAME</workspaceName>
<projectName>PROJECT_NAME</projectName>
<scenarioName>SCENARIO_NAME</scenarioName>
```
The `API_KEY` is available [on your profile page](https://api.octoperf.com/doc/account/profile/#apikey). The other values are the textual names of each element inside OctoPerf.

### Result

Before launch it should look like this inside your Azure pipeline:

![pipeline](https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/ci/img/pipeline.png)

And when you execute it:

![azure-devops](https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/ci/img/azure-devops.png)

Congratulations! You've just triggered a load test on OctoPerf  using **Azure Devops** and **OctoPerf's Maven Plugin**.

![GitLab Build Running On OctoPerf](https://api.octoperf.com/doc/runtime/edit-scenario/integrations-automation/ci-cd/ci/img/gitlab-analysis-report.png)

The analysis report is visible in OctoPerf too. For further information, see our [OctoPerf Maven Plugin](https://github.com/OctoPerf/octoperf-maven-plugin) documentation.