Version: 1.8.0

Continuous Integration

Playwright tests can be executed in CI environments. We have created sample configurations for common CI providers.

Introduction#

3 steps to get your tests running on CI:

  1. Ensure CI agent can run browsers: Use our Docker image in Linux agents. Windows and macOS agents do not require any additional dependencies.

  2. Install Playwright:

    $ pip install playwright==1.8.0a1
    $ playwright install

  3. Run your tests:

    $ pytest

CI configurations#

GitHub Actions#

The Playwright GitHub Action can be used to run Playwright tests on GitHub Actions.

steps:
- uses: microsoft/playwright-github-action@v1
- name: Set up Python
uses: actions/setup-python@v2
with:
python-version: 3.8
- name: Install dependencies
run: |
python -m pip install --upgrade pip
pip install playwright==1.8.0a1
pip install -e .
- name: Ensure browsers are installed
run: python -m playwright install
- name: Run your tests
run: pytest

We run our tests on GitHub Actions, across a matrix of 3 platforms (Windows, Linux, macOS) and 3 browsers (Chromium, Firefox, WebKit).

Docker#

We have a pre-built Docker image which can either be used directly, or as a reference to update your existing Docker definitions.

Suggested configuration

  1. By default, Docker runs a container with a /dev/shm shared memory space 64MB. This is typically too small for Chromium and will cause Chromium to crash when rendering large pages. To fix, run the container with docker run --shm-size=1gb to increase the size of /dev/shm. Since Chromium 65, this is no longer necessary. Instead, launch the browser with the --disable-dev-shm-usage flag:

    browser = playwright.chromium.launch({
    args=['--disable-dev-shm-usage']
    })

    This will write shared memory files into /tmp instead of /dev/shm. See crbug.com/736452 for more details.

  2. Using --ipc=host is also recommended when using Chromium—without it Chromium can run out of memory and crash. Learn more about this option in Docker docs.

  3. Seeing other weird errors when launching Chromium? Try running your container with docker run --cap-add=SYS_ADMIN when developing locally.

  4. dumb-init is worth checking out if you're experiencing a lot of zombies Chromium processes sticking around. There's special treatment for processes with PID=1, which makes it hard to terminate Chromium properly in some cases (e.g. in Docker).

Azure Pipelines#

For Windows or macOS agents, no additional configuration required, just install Playwright and run your tests.

For Linux agents, you can use our Docker container with Azure Pipelines support for running containerized jobs. Alternatively, you can refer to the Dockerfile to see additional dependencies that need to be installed on a Ubuntu agent.

pool:
vmImage: 'ubuntu-18.04'
container: mcr.microsoft.com/playwright:bionic
steps:
...

Travis CI#

Suggested configuration

  1. User namespace cloning should be enabled to support proper sandboxing
  2. xvfb should be launched in order to run Chromium in non-headless mode (e.g. to test Chrome Extensions)
  3. If your project does not have package-lock.json, Travis would be auto-caching node_modules directory. If you run npm install (instead of npm ci), it is possible that the browser binaries are not downloaded. Fix this with these steps outlined below.

To sum up, your .travis.yml might look like this:

language: node_js
dist: bionic
addons:
apt:
packages:
# These are required to run webkit
- libwoff1
- libopus0
- libwebp6
- libwebpdemux2
- libenchant1c2a
- libgudev-1.0-0
- libsecret-1-0
- libhyphen0
- libgdk-pixbuf2.0-0
- libegl1
- libgles2
- libevent-2.1-6
- libnotify4
- libxslt1.1
- libvpx5
# gstreamer and plugins to support video playback in WebKit.
- gstreamer1.0-gl
- gstreamer1.0-plugins-base
- gstreamer1.0-plugins-good
- gstreamer1.0-plugins-bad
# This is required to run chromium
- libgbm1
# this is needed for running headful tests
- xvfb
# allow headful tests
before_install:
# Enable user namespace cloning
- "sysctl kernel.unprivileged_userns_clone=1"
# Launch XVFB
- "export DISPLAY=:99.0"
- "sh -e /etc/init.d/xvfb start"

CircleCI#

Running Playwright on CircleCI requires the following steps:

  1. Use the pre-built Docker image in your config like so:

    docker:
    - image: mcr.microsoft.com/playwright:bionic
    environment:
    NODE_ENV: development # Needed if playwright is in `devDependencies`

  2. If you’re using Playwright through Jest, then you may encounter an error spawning child processes:

    [00:00.0] jest args: --e2e --spec --max-workers=36
    Error: spawn ENOMEM
    at ChildProcess.spawn (internal/child_process.js:394:11)

    This is likely caused by Jest autodetecting the number of processes on the entire machine (36) rather than the number allowed to your container (2). To fix this, set jest --maxWorkers=2 in your test command.

Jenkins#

Jenkins supports Docker agents for pipelines. Use the Playwright Docker image to run tests on Jenkins.

pipeline {
agent { docker { image 'mcr.microsoft.com/playwright:bionic' } }
stages {
stage('e2e-tests') {
steps {
sh 'npm install'
sh 'npm run test'
}
}
}
}

Bitbucket Pipelines#

Bitbucket Pipelines can use public Docker images as build environments. To run Playwright tests on Bitbucket, use our public Docker image (see Dockerfile).

image: mcr.microsoft.com/playwright:bionic

While the Docker image supports sandboxing for Chromium, it does not work in the Bitbucket Pipelines environment. To launch Chromium on Bitbucket Pipelines, use the chromiumSandbox: false launch argument.

browser = playwright.chromium.launch(chromiumSandbox=False)

GitLab CI#

To run Playwright tests on GitLab, use our public Docker image (see Dockerfile).

stages:
- test
tests:
stage: test
image: mcr.microsoft.com/playwright:bionic
script:
...

Caching browsers#

By default, Playwright downloads browser binaries when the Playwright NPM package is installed. The NPM packages have a postinstall hook that downloads the browser binaries. This behavior can be customized with environment variables.

Caching browsers on CI is strictly optional: The postinstall hooks should execute and download the browser binaries on every run.

Exception: node_modules are cached (Node-specific)#

Most CI providers cache the npm-cache directory (located at $HOME/.npm). If your CI pipelines caches the node_modules directory and you run npm install (instead of npm ci), the default configuration

will not work. This is because the npm install step will find the Playwright NPM package on disk and not execute the postinstall step.

Travis CI automatically caches node_modules if your repo does not have a package-lock.json file.

This behavior can be fixed with one of the following approaches:

  1. Move to caching $HOME/.npm or the npm-cache directory. (This is the default behavior in most CI providers.)
  2. Set PLAYWRIGHT_BROWSERS_PATH=0 as the environment variable before running npm install. This will download the browser binaries in the node_modules directory and cache them with the package code. See installation docs.
  3. Use npm ci (instead of npm install) which forces a clean install: by removing the existing node_modules directory. See npm docs.
  4. Cache the browser binaries, with the steps below.

Directories to cache#

With the default behavior, Playwright downloads the browser binaries in the following directories:

  • %USERPROFILE%\AppData\Local\ms-playwright on Windows
  • ~/Library/Caches/ms-playwright on MacOS
  • ~/.cache/ms-playwright on Linux

To cache the browser downloads between CI runs, cache this location in your CI configuration, against a hash of the Playwright version.

Debugging browser launches#

Playwright supports the DEBUG environment variable to output debug logs during execution. Setting it to pw:browser* is helpful while debugging Error: Failed to launch browser errors.

DEBUG=pw:browser* pytest

Running headful#

By default, Playwright launches browsers in headless mode. This can be changed by passing a flag when the browser is launched.

from playwright.sync_api import sync_playwright
with sync_playwright() as p:
# Works across chromium, firefox and webkit
browser = p.chromium.launch(headless=False)

On Linux agents, headful execution requires Xvfb to be installed. Our Docker image and GitHub Action have Xvfb pre-installed. To run browsers in headful mode with Xvfb, add xvfb-run before the Node.js command.

xvfb-run python test.py