119.md 9.2 KB
Newer Older
Lab机器人's avatar
readme  
Lab机器人 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
# Load Performance Testing

> 原文:[https://docs.gitlab.com/ee/user/project/merge_requests/load_performance_testing.html](https://docs.gitlab.com/ee/user/project/merge_requests/load_performance_testing.html)

*   [How Load Performance Testing works](#how-load-performance-testing-works)
*   [Configure the Load Performance Testing job](#configure-the-load-performance-testing-job)
    *   [Determine the test parameters](#determine-the-test-parameters)
    *   [Test Environment setup](#test-environment-setup)
    *   [Write the load performance test](#write-the-load-performance-test)
    *   [Configure the test in GitLab CI/CD](#configure-the-test-in-gitlab-cicd)
    *   [Load Performance testing in Review Apps](#load-performance-testing-in-review-apps)

# Load Performance Testing[](#load-performance-testing-premium "Permalink")

[Introduced](https://gitlab.com/gitlab-org/gitlab/-/issues/10683) in [GitLab Premium](https://about.gitlab.com/pricing/) 13.2.

使用负载性能测试,您可以在[GitLab CI / CD 中](../../../ci/README.html)测试任何待处理代码更改对应用程序后端的影响.

GitLab 使用免费的开源工具[k6](https://k6.io/)来测量负载下应用程序的系统性能.

Unlike [Browser Performance Testing](browser_performance_testing.html), which is used to measure how web sites perform in client browsers, Load Performance Testing can be used to perform various types of [load tests](https://k6.io/docs/#use-cases) against application endpoints such as APIs, Web Controllers, and so on. This can be used to test how the backend or the server performs at scale.

例如,您可以使用负载性能测试对应用程序中流行的 API 端点执行许多并发 GET 调用,以查看其性能.

## How Load Performance Testing works[](#how-load-performance-testing-works "Permalink")

首先,在您的`.gitlab-ci.yml`文件中定义一个作业,以生成" [负载性能"报告工件](../../../ci/pipelines/job_artifacts.html#artifactsreportsload_performance-premium) . GitLab 会检查此报告,比较源分支和目标分支之间的关键负载性能指标,然后在合并请求小部件中显示信息:

[![Load Performance Widget](img/1d9bd6ef20c22f103a6d120b8c673b06.png)](img/load_performance_testing.png)

接下来,您需要配置测试环境并编写 k6 测试.

测试完成后,合并请求窗口小部件显示的关键性能指标是:

*   支票:在 k6 测试中配置的[支票的](https://k6.io/docs/using-k6/checks)通过率.
*   TTFB P90: The 90th percentile of how long it took to start receiving responses, aka the [Time to First Byte](https://en.wikipedia.org/wiki/Time_to_first_byte) (TTFB).
*   TTFB P95:TTFB 的第 95 个百分点.
*   RPS:测试能够达到的平均每秒请求数(RPS).

**注意:**如果"负载性能"报告没有可比较的数据,例如,当您第一次在`.gitlab-ci.yml`添加"负载性能"作业时,"负载性能"报告小部件将不会显示. 它必须在目标分支(例如`master` )上至少运行一次,然后才能在针对该分支的合并请求中显示.

## Configure the Load Performance Testing job[](#configure-the-load-performance-testing-job "Permalink")

配置负载性能测试作业可以分为几个不同的部分:

*   确定测试参数,例如吞吐量,等等.
*   设置目标测试环境以进行负载性能测试.
*   设计并编写 k6 测试.

### Determine the test parameters[](#determine-the-test-parameters "Permalink")

您需要做的第一件事是确定要运行[的负载测试](https://k6.io/docs/test-types/introduction)[类型](https://k6.io/docs/test-types/introduction)以及运行方式(例如,用户数量,吞吐量等).

请参阅[k6 文档](https://k6.io/docs/) ,尤其是[k6 测试指南](https://k6.io/docs/testing-guides) ,以获取有关上述内容及更多内容的指南.

### Test Environment setup[](#test-environment-setup "Permalink")

负载性能测试的大部分工作是为高负载准备目标测试环境. 您应该确保它能够处理将要测试的[吞吐量](https://k6.io/blog/monthly-visits-concurrent-users) .

通常还需要在目标环境中具有代表性的测试数据,以供负载性能测试使用.

强烈建议[不要在生产环境中运行这些测试](https://k6.io/our-beliefs#load-test-in-a-pre-production-environment) .

### Write the load performance test[](#write-the-load-performance-test "Permalink")

准备好环境后,您可以编写 k6 测试本身. k6 是一种灵活的工具,可用于运行[多种性能测试](https://k6.io/docs/test-types/introduction) . 有关如何编写测试的详细信息,请参阅[k6 文档](https://k6.io/docs/) .

### Configure the test in GitLab CI/CD[](#configure-the-test-in-gitlab-cicd "Permalink")

准备好 k6 测试后,下一步就是在 GitLab CI / CD 中配置负载性能测试作业. 最简单的方法是使用 GitLab 随附的[`Verify/Load-Performance-Testing.gitlab-ci.yml`](https://gitlab.com/gitlab-org/gitlab/blob/master/lib/gitlab/ci/templates/Verify/Load-Performance-Testing.gitlab-ci.yml)模板.

**注意:**对于大型 k6 测试,您需要确保执行实际测试的 GitLab Runner 实例能够处理运行测试. 有关规格的详细信息,请参阅[k6 的指南](https://k6.io/docs/testing-guides/running-large-tests#hardware-considerations) . [默认的共享 GitLab.com 运行程序](../../gitlab_com/#linux-shared-runners)可能没有足够的规格来处理大多数大型 k6 测试.

该模板在作业中运行[k6 Docker 容器](https://hub.docker.com/r/loadimpact/k6/) ,并提供了几种自定义作业的方法.

配置工作流程示例:

1.  设置一个可以运行 Docker 容器的 GitLab Runner,例如使用[Docker-in-Docker 工作流](../../../ci/docker/using_docker_build.html#use-docker-in-docker-workflow-with-docker-executor)的 Runner.
2.`.gitlab-ci.yml`文件中配置默认​​的 Load Performance Testing CI 作业. 您需要包括模板并使用变量进行配置:

    ```
    include:
      template: Verify/Load-Performance-Testing.gitlab-ci.yml

    load_performance:
      variables:
        K6_TEST_FILE: <PATH TO K6 TEST FILE IN PROJECT> 
    ```

上面的示例在运行 k6 测试的 CI / CD 管道中创建了一个`load_performance`作业.

**注意:**对于 Kubernetes 设置,应使用其他模板: [`Jobs/Load-Performance-Testing.gitlab-ci.yml`](https://gitlab.com/gitlab-org/gitlab/blob/master/lib/gitlab/ci/templates/Jobs/Load-Performance-Testing.gitlab-ci.yml) .

k6 具有[各种选项](https://k6.io/docs/using-k6/options)来配置它将如何运行测试,例如运行什么吞吐量(RPS),测试应该运行多长时间等等. 几乎所有选项都可以在测试本身中进行配置,但是您也可以通过`K6_OPTIONS`变量传递命令行选项.

例如,您可以使用 CLI 选项覆盖测试的持续时间:

```
 include:
    template: Verify/Load-Performance-Testing.gitlab-ci.yml

  load_performance:
    variables:
      K6_TEST_FILE: <PATH TO K6 TEST FILE IN PROJECT>
      K6_OPTIONS: '--duration  30s' 
```

如果 k6 的结果通过[摘要导出](https://k6.io/docs/results-visualization/json#summary-export)保存为" [负载性能"报告工件,则](../../../ci/pipelines/job_artifacts.html#artifactsreportsload_performance-premium) GitLab 仅在 MR 小部件中显示关键性能指标. 始终使用最新的可用负载性能工件.

如果启用了[GitLab Pages,](../pages/index.html)则可以直接在浏览器中查看报告.

### Load Performance testing in Review Apps[](#load-performance-testing-in-review-apps "Permalink")

上面的 CI / CD YAML 配置示例适用于针对静态环境进行测试,但是可以通过一些额外的步骤将其扩展为与[审阅应用程序](../../../ci/review_apps)[动态环境](../../../ci/environments)一起使用.

最好的方法是将动态 URL 捕获到一个自定义环境变量中,然后由`load_performance`作业[继承](../../../ci/variables/README.html#inherit-environment-variables)`load_performance` . 然后应将要运行的 k6 测试脚本配置为使用该环境 URL,例如: `http.get(`${__ENV.ENVIRONMENT_URL`})` .

例如:

1.`review`工作中:
    1.  捕获动态 URL 并将其保存到`.env`文件,例如`echo "ENVIRONMENT_URL=$CI_ENVIRONMENT_URL" >> review.env` .
    2.`.env`文件设置为[`artifacts:reports:dotenv` report](../../../ci/variables/README.html#inherit-environment-variables) .
2.`load_performance`作业设置为依赖于审阅作业,以便它继承环境变量.
3.  配置 k6 测试脚本以在其步骤中使用环境变量.

您的`.gitlab-ci.yml`文件可能类似于:

```
stages:
  - deploy
  - performance

include:
  template: Verify/Load-Performance-Testing.gitlab-ci.yml

review:
  stage: deploy
  environment:
    name: review/$CI_COMMIT_REF_NAME
    url: http://$CI_ENVIRONMENT_SLUG.example.com
  script:
    - run_deploy_script
    - echo "ENVIRONMENT_URL=$CI_ENVIRONMENT_URL" >> review.env
  artifacts:
    reports:
      dotenv:
        review.env
  rules:
    - if: '$CI_COMMIT_BRANCH'  # Modify to match your pipeline rules, or use `only/except` if needed.

load_performance:
  dependencies:
    - review
  rules:
    - if: '$CI_COMMIT_BRANCH'  # Modify to match your pipeline rules, or use `only/except` if needed. 
```