[英]How to run docker-compose inside docker in docker which runs inside gitlab-runner container?
I have a gitlab runner inside a docker container, runs fine if I run an image like nginx.我在 docker 容器中有一个 gitlab runner,如果我运行像 nginx 这样的图像,运行良好。 But now I tried to run docker in docker (dind) inside the gitlab runner and I want to run docker-compose inside the dind.但是现在我尝试在 gitlab runner 中的 docker (dind) 中运行 docker,我想在 dind 中运行 docker-compose。 Docker info runs fine, but if I try to run docker-compose I get an permission-denied error. Docker 信息运行良好,但如果我尝试运行 docker-compose,我会收到一个权限被拒绝的错误。
I linked the /usr/local/bin/docker-compose file to the gitlab runner container and enter it in the volumes parameter in the runner config.toml file.我将 /usr/local/bin/docker-compose 文件链接到 gitlab runner 容器,并将其输入到 runner config.toml 文件的卷参数中。
If I try to run sudo it ends with an unknown command error, so that could not be the solution.如果我尝试运行 sudo,它会以未知的命令错误结束,因此这不是解决方案。
Do I have to link some file more or are the to many nested containers?我是否必须更多地链接一些文件,还是链接到许多嵌套容器?
In order to have docker-compose
, you need to install it for image docker , which is of version 18.09.6, build 481bc77 at the time of writing.为了让docker-compose
,您需要安装它的图像泊坞窗,这在写作的时候是18.09.6版本,构建481bc77。
Since docker-compose version 1.24.0 , you also need the following dependencies for docker-compose to be installed on alpine :从docker-compose 版本 1.24.0 开始,您还需要以下依赖项才能将docker -compose 安装在 alpine 上:
apk add py-pip python3-dev libffi-dev openssl-dev gcc libc-dev make
Here is a sample .gitlab-ci.yml
:这是一个示例.gitlab-ci.yml
:
image: docker:stable
stages:
- deploy
services:
- docker:dind
before_script:
- apk update
- apk add py-pip python3-dev libffi-dev openssl-dev gcc libc-dev make
- pip install docker-compose
deploy_app:
stage: deploy
script:
- docker-compose down
- docker-compose up -d
if you are using dind it means docker is working OK, now you just have to install docker-compose that is just simple python package and you can do it in before_script如果您使用的是 dind,则意味着 docker 工作正常,现在您只需要安装 docker-compose 这只是简单的 python 包,您可以在 before_script 中完成
.gitlab-ci.yml .gitlab-ci.yml
image: docker:latest
services:
- docker:dind
variables:
DOCKER_DRIVER: overlay2
stages:
- test
before_script:
- apk add --no-cache py-pip
- pip install docker-compose
- docker info
- docker login -u gitlab-ci-token -p $CI_JOB_TOKEN docker.registry.com
test:
stage: test
script:
- cp .env.sample .env # copy environement variable
- docker-compose up -d
# run some test here
The best choice for running docker-compose inside docker in docker is using the docker/compose image as bellow:在 docker 中运行 docker-compose 的最佳选择是使用docker/compose图像,如下所示:
job-x:
image: docker/compose
script:
- docker-compose version
A real and full example of this is using docker-compose to deploy with specific installed runner on a server as bellow:一个真实且完整的示例是使用 docker-compose 在服务器上部署特定安装的运行器,如下所示:
image: docker:stable
services:
- docker:dind
stages:
- build
- deploy
before_script:
# resolve TAG
# Default branch leaves tag latest
# All other branches are tagged with the escaped branch name (commit ref slug)
- |
if [[ "$CI_COMMIT_BRANCH" == "$CI_DEFAULT_BRANCH" ]]; then
TAG="latest"
echo "Running on default branch '$CI_DEFAULT_BRANCH': tag = '$TAG'"
else
TAG="$CI_COMMIT_REF_SLUG"
echo "Running on branch '$CI_COMMIT_BRANCH': tag = $TAG"
fi
docker-build:
stage: build
script:
- echo "user ${CI_REGISTRY_USER}"
- echo "registry ${CI_REGISTRY}"
- docker login -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $CI_REGISTRY
- docker build --build-arg BUILDKIT_INLINE_CACHE=1 --cache-from "$CI_REGISTRY_IMAGE:${TAG}" -t "$CI_REGISTRY_IMAGE:${TAG}" .
- docker push "$CI_REGISTRY_IMAGE:${TAG}"
when: manual
deploy-staging:
image: docker/compose
stage: deploy
variables:
IMAGE: $CI_REGISTRY_IMAGE
TAG: $TAG
HOST_PORT: $HOST_PORT
script:
- docker-compose version
- docker-compose -f docker-compose.yml up -d --no-build
when: manual
The docker-compose.yml file seems like: docker-compose.yml 文件看起来像:
version: "3.7"
services:
web:
image: $IMAGE:$TAG
build:
context: .
ports:
- $HOST_PORT:8000
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.