优化 Docker 镜像构建和缓存的 GitHub Action
docker-build-with-cache-action 是一个 GitHub Action,用于构建 Docker 镜像并缓存构建阶段,提高后续构建速度。它支持多阶段构建和多个主流镜像仓库,如 Docker Hub、AWS ECR、GitHub 和 Google Cloud 等。该 Action 可推送镜像和缓存阶段,实现增量构建。此外,它兼容 Docker Compose 文件,提供自定义标签、构建参数和 Dockerfile 路径等灵活配置。通过优化构建流程,简化了 Docker 镜像的构建和发布过程,提升 CI/CD 效率,适用于各种持续集成场景。
This action builds your docker image and caches the stages (supports multi-stage builds) to improve building times in subsequent builds.
By default, it pushes the image with all the stages to a registry (needs username and password), but you can disable this feature by setting push_image_and_stages
to false
.
Built-in support for the most known registries:
Docker updated to 24.0.6
BuildKit is enabled for faster/parallel builds
Cache also works with BuildKit enabled except for old GitHub Docker Registry (docker.pkg.github.com). You can either migrate to ghcr.io or disable BuildKit to use the old registry:
- name: Build with DOCKER_BUILDKIT disabled for old GitHub Docker Registry
uses: whoan/docker-build-with-cache-action@master
env:
DOCKER_BUILDKIT: 0
with:
registry: docker.pkg.github.com
...
or
:star2: New in v5.10.0: Now you can use overrides for your compose file(s) like this:
docker-compose.yml > docker-compose.override.yml > docker-compose.override2.yml
image_tag: Tag(s) of the image. Allows multiple comma-separated tags (e.g. one,another
) (default: latest
).
If you set compose_file and the image(s) already has/have a tag, this is ignored.
context: Docker context (default: ./
). If a compose_file is provided, it will be the context prefixed to any additional context read from the compose file. Look at #133 for more details.
registry: Docker registry (default: Docker Hub's registry). You need a registry to use the cache functionality.
username: Docker registry's user (needed to push images, or to pull from a private repository).
password: Docker registry's password (needed to push images, or to pull from a private repository).
session: Extra auth parameters. For AWS ECR, means setting AWS_SESSION_TOKEN environment variable.
push_git_tag: In addition to image_tag
, you can also push the git tag in your branch tip (default: false
).
pull_image_and_stages: Set to false
to avoid pulling from the registry or to build from scratch (default: true
).
stages_image_name: Set custom name for stages/cache image (default: ${image_name}-stages
). Tags are ignored.
push_image_and_stages: Test a command before pushing. Use false
to not push at all (default: true
).
This input also supports 2 special values, which are useful if your workflow can be triggered by different events:
on:push
: Push only if the workflow was triggered by a push.on:pull_request
: Push only if the workflow was triggered by a pull_request.services_regex: Regex to filter services from compose file. Only valid when compose_file was provided. Default is .+
(all services).
compose_file
is setimage_name
dockerfile: Dockerfile filename path (default: "$context"/Dockerfile
).
build_extra_args: Extra params for docker build
(e.g. "--build-arg=hello=world"
).
:star2: New in v5.11.0: If you need extra args with newlines or spaces, use json format like this:
build_extra_args: '{"--build-arg": "myarg=Hello\nWorld"}'
:star2: If you need multiple args with same key, use an array as the value of the key like this:
build_extra_args: '{"--build-arg": ["foo=bar", "one=two"]}'
docker.pkg.github.com/whoan/hello-world/hello-world
The action does the following every time it is triggered:
registry
(default: https://hub.docker.com)image_tag
<image_name>-stages:<1,2,3,...>
<image_name>:<git_tag>
if you set push_git_tag: true
Find working minimal examples for the most known registries in this repo.
If you don't specify a registry, Docker Hub is the default one
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.DOCKER_HUB_PASSWORD }}" image_name: hello-world
GitHub automatically creates a GITHUB_TOKEN secret to use in your workflow. If you are going to use the new GitHub Registry (ghcr.io), be sure to use a Personal Access Token (as the password) with "write:packages" and "read:packages" scopes. More info here.
If you push the image to a public repository's GitHub Registry, please be aware that it will be impossible to delete it because of GitHub's policy (see Deleting a package).
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.GITHUB_TOKEN }}" registry: docker.pkg.github.com #or #registry: ghcr.io image_name: hello-world
More info here on how to get GCloud JSON key.
- uses: whoan/docker-build-with-cache-action@v5 with: username: _json_key password: "${{ secrets.GCLOUD_JSON_KEY }}" registry: gcr.io image_name: hello-world
You don't even need to create the repositories in advance, as this action takes care of that for you! (you'll need the
CreateRepository
permission)
- uses: whoan/docker-build-with-cache-action@v5 with: username: "${{ secrets.AWS_ACCESS_KEY_ID }}" # no need to provide it if you already logged in with aws-actions/configure-aws-credentials password: "${{ secrets.AWS_SECRET_ACCESS_KEY }}" # no need to provide it if you already logged in with aws-actions/configure-aws-credentials session: "${{ secrets.AWS_SESSION_TOKEN }}" # if you need role assumption. no need to provide it if you already logged in with aws-actions/configure-aws-credentials # private registry registry: 861729690598.dkr.ecr.us-west-1.amazonaws.com # or public registry #registry: public.ecr.aws image_name: hello-world
The compose file is parsed and the action will run once for each detected image. The registry is also detected from the image name, and if none is provided, DockerHub is assumed.
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.DOCKER_HUB_PASSWORD }}" compose_file: docker-compose.yml
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.GITHUB_TOKEN }}" registry: docker.pkg.github.com compose_file: docker-compose.yml
With a compose file override:
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.DOCKER_HUB_PASSWORD }}" compose_file: docker-compose.yml > docker-compose.override.yml
Filtering services by regex:
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.GITHUB_TOKEN }}" registry: docker.pkg.github.com compose_file: docker-compose.yml services_regex: '(service_1|extra_service.*)' # eg: builds services called exactly "service_1" plus the ones which start with "extra_service" and may have extra chars after
- uses: whoan/docker-build-with-cache-action@v5 with: username: whoan password: "${{ secrets.GITHUB_TOKEN }}" image_name: whoan/docker-images/node image_tag: alpine-slim,another-tag,latest push_git_tag: true registry: docker.pkg.github.com context: node-alpine-slim dockerfile: custom.Dockerfile build_extra_args: "--compress=true --build-arg=hello=world" push_image_and_stages: docker run my_awesome_image:latest # eg: push only if docker run succeed
latest
tag, it may download different versions when the action is triggered, and it will invalidate the cache.The tests for this action are run in a separate repo as I need to set credentials for each registry with GitHub secrets and doing so in this repo is not practical.
AI辅助编程,代码自动修复
Trae是一种自适应的集成开发环境(IDE),通过自动化和多元协作改变开发流程。利用Trae,团队能够更快速、精确地编写和部署代码,从而提高编程效率和项目交付速度。Trae具备上下文感知和代码自动完成功能,是提升开发效率的理想工具。
AI小说写作助手,一站式润色、改写、扩写
蛙蛙写作—国内先进的AI写作平台,涵盖小说、学术、社交媒体等多场景。提供续写、改写、润色等功能,助力创作者高效优化写作流程。界面简洁,功能全面,适合各类写作者提升内容品质和工作效率。
全能AI智能助手,随时解答生活与工作的多样问题
问小白,由元石科技研发的AI智能助手,快速准确地解答各种生活和工作问题,包括但不限于搜索、规划和社交互动,帮助用户在日常生活中提高效率,轻松管理个人事务。
实时语音翻译/同声传译工具
Transly是一个多场景的AI大语言模型驱动的同声传译、专业翻译助手,它拥有超精准的音频识别翻译能力,几乎零延迟的使用体验和支持多国语言可以让你带它走遍全球,无论你是留学生、商务人士、韩剧美剧爱好者,还是出国游玩、多国会议、跨国追星等等,都可以满足你所有需要同传的场景需求,线上线下通用,扫除语言障碍,让全世界的语言交流不再有国界。
一键生成PPT和Word,让学习生活更轻松
讯飞智文是一个利用 AI 技术的项目,能够帮助用户生成 PPT 以及各类文档。无论是商业领域的市场分析报告、年度目标制定,还是学生群体的职业生涯规划、实习避坑指南,亦或是活动策划、旅游攻略等内容,它都能提供支持,帮助用户精准表达,轻松呈现各种信息。
深度推理能力全新升级,全面对标OpenAI o1
科大讯飞的星火大模型,支持语言理解、知识问答和文本创作等多功能,适用于多种文件和业务场景,提升办公和日常生活的效率。讯飞星火是一个提供丰富智能服务的平台,涵盖科技资讯、图像创作、写作辅助、编程解答、科研文献解读等功能,能为不同需求的用户提供便捷高效的帮助,助力用户轻松获取信息、解决问题,满足多样化使用场景。
一种基于大语言模型的高效单流解耦语音令牌文本到语音合成模型
Spark-TTS 是一个基于 PyTorch 的开源文本到语音合成项目,由多个知名机构联合参与。该项目提供了高效的 LLM(大语言模型)驱动的语音合成方案,支持语音克隆和语音创建功能,可通过命令行界面(CLI)和 Web UI 两种方式使用。用户可以根据需求调整语音的性别、音高、速度等参数,生成高质量的语音。该项目适用于多种场景,如有声读物制作、智能语音助手开发等。
AI助力,做PPT更简单!
咔片是一款轻量化在线演示设计工具,借助 AI 技术,实现从内容生成到智能设计的一站式 PPT 制作服务。支持多种文档格式导入生成 PPT,提供海量模板、智能美化、素材替换等功能,适用于销售、教师、学生等各类人群,能高效制作出高品质 PPT,满足不同场景演示需求。
选题、配图、成文,一站式创作,让内容运营更高效
讯飞绘文,一个AI集成平台,支持写作、选题、配图、排版和发布。高效生成适用于各类媒体的定制内容,加速品牌传播,提升内容营销效果。
专业的AI公文写作平台,公文写作神器
AI 材料星,专业的 AI 公文写作辅助平台,为体制内工作人员提供高效的公文写作解决方案。拥有海量公文文库、9 大核心 AI 功能,支持 30 + 文稿类型生成,助力快速完成领导讲话、工作总结、述职报告等材料,提升办公效率,是体制打工人的得力写作神器。
最新AI工具、AI资讯
独家AI资源、AI项目落地
微信扫一扫关注公众号