[buildx] buildx 사용하기 - volume 마운트 실행

사용

buildx builder 컨테이너를 생성할 때, 아래 명령어를 이용하여 생성할 수 있다.

docker buildx create --name builder --use --bootstrap

위의 명령어를 사용하여 생성하면 원하는 vloume을 마운트 하는게 불가능해서 약간 편법을 이용하기로 한다. 먼저 아래의 내용을 builder라는 이름으로 $HOME/.docker/buildx/instances 경로에 저장한다.

{"Name":"builder","Driver":"docker-container","Nodes":[{"Name":"builder0","Endpoint":"unix:///var/run/docker.sock","Platforms":null,"Flags":null,"DriverOpts":null,"Files":null}],"Dynamic":false}

그리고 buildx 컨테이너를 생성한다.

mkdir $HOME/volume
docker run -d -v $HOME/volume:/var/lib/buildkit --privileged --name buildx_buildkit_builder0 moby/buildkit:buildx-stable-1

컨테이너가 생성되면, 아래 명령어로 buildx builder를 기본 builder로 지정한다.

docker buildx use builder

docker buildx ls
NAME/NODE  DRIVER/ENDPOINT             STATUS  BUILDKIT PLATFORMS
builder *  docker-container
  builder0 unix:///var/run/docker.sock running v0.10.5  linux/amd64, linux/amd64/v2, linux/amd64/v3, linux/amd64/v4, linux/386
default    docker
  default  default                     running 20.10.17 linux/amd64, linux/386

---

volume 보존 테스트

위에서 $HOME/volume 디렉토리를 마운트한 뒤, builder 컨테이너를 삭제한 뒤에 다시 생성해도 빌드캐시가 유지되는지 테스트 해본다. 먼저 Dockerfile을 임시로 빌드한다. Dockerfile 내용은 아래와 같다.

FROM python:3.7
ENV PYTHONUNBUFFERED 1
ENV PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION python
RUN mkdir /home/ubuntu
WORKDIR /home/ubuntu

RUN apt-get update && apt-get install -y libgl1-mesa-dev
RUN pip install Tensorflow==2.2
RUN pip install opencv-python==4.5.1.48

Dockerfile을 builder를 이용해서 빌드한다.

docker buildx build .
WARNING: No output specified with docker-container driver. Build result will only remain in the build cache. To push result image into registry use --push or to load image into docker use --load      
[+] Building 237.1s (9/9) FINISHED
 => [internal] load build definition from Dockerfile                                           0.1s 
 => => transferring dockerfile: 361B                                                           0.0s 
 => [internal] load .dockerignore                                                              0.1s 
 => => transferring context: 2B                                                                0.0s 
 => [internal] load metadata for docker.io/library/python:3.7                                  3.4s 
 => [1/6] FROM docker.io/library/python:3.7@sha256:bf85a74f4ace82f3503c2199aaae10e7a8e370bc7  60.7s 
 => => resolve docker.io/library/python:3.7@sha256:bf85a74f4ace82f3503c2199aaae10e7a8e370bc7e  0.0s 
 => => sha256:9a5f30b220ac2d88165d2e71dec428a6846050e24b4a8cedb8c1d90b46780cc 2.89MB / 2.89MB  0.7s 
 => => sha256:24e64f2eb9381ce66436fa7d151419d087b854b218c6abc2e0e94915c00fc 15.48MB / 15.48MB  3.8s 
 => => sha256:ece0ca00f21cd46fbcf08f8172c1b8d9be5d9c68f3a53e107279a31ee30bba2e 234B / 234B     0.5s 
 => => sha256:0f67f32c26d393a2580062f2cebfde80cc4c5a5e264bbb7a32569c6c7551c1c 6.29MB / 6.29MB  4.3s 
 => => sha256:60b38700e7fb2cdfac79b15e4c1691a80fe6b4101c7b7fea66b9e7cd64 196.88MB / 196.88MB  35.5s 
 => => sha256:81283a9569ad5e90773f038daedd0d565810ca5935eec8f53b8bcb6a1990 54.58MB / 54.58MB  21.4s 
 => => sha256:c796299bbbddc7aeada9539a4e7874a75fa2b6ff421f8d5ad40f227b40ab4 10.88MB / 10.88MB  3.6s 
 => => sha256:fa1d4c8d85a4e064e50cea74d4aa848dc5fc275aef223fcc1f21fbdb1b5dd18 5.16MB / 5.16MB  2.3s 
 => => sha256:32de3c850997ce03b6ff4ae8fb00b34b9d7d7f9a35bfcdb8538e22cc7b77 55.03MB / 55.03MB  16.4s 
 => => extracting sha256:32de3c850997ce03b6ff4ae8fb00b34b9d7d7f9a35bfcdb8538e22cc7b77c29d      1.8s 
 => => extracting sha256:fa1d4c8d85a4e064e50cea74d4aa848dc5fc275aef223fcc1f21fbdb1b5dd182      0.2s 
 => => extracting sha256:c796299bbbddc7aeada9539a4e7874a75fa2b6ff421f8d5ad40f227b40ab4d86      0.2s 
 => => extracting sha256:81283a9569ad5e90773f038daedd0d565810ca5935eec8f53b8bcb6a199030d6      6.6s 
 => => extracting sha256:60b38700e7fb2cdfac79b15e4c1691a80fe6b4101c7b7fea66b9e7cd64d961cf     12.5s 
 => => extracting sha256:0f67f32c26d393a2580062f2cebfde80cc4c5a5e264bbb7a32569c6c7551c1c2     11.2s 
 => => extracting sha256:24e64f2eb9381ce66436fa7d151419d087b854b218c6abc2e0e94915c00fc4bd      0.5s 
 => => extracting sha256:ece0ca00f21cd46fbcf08f8172c1b8d9be5d9c68f3a53e107279a31ee30bba2e      0.1s 
 => => extracting sha256:9a5f30b220ac2d88165d2e71dec428a6846050e24b4a8cedb8c1d90b46780cce      0.2s 
 => [2/6] RUN mkdir /home/ubuntu                                                               3.0s 
 => [3/6] WORKDIR /home/ubuntu                                                                 0.2s 
 => [4/6] RUN apt-get update && apt-get install -y libgl1-mesa-dev                            18.8s 
 => [5/6] RUN pip install Tensorflow==2.2                                                    138.6s 
 => [6/6] RUN pip install opencv-python==4.5.1.48                                             12.2s

이 상태에서 builder를 제거한 뒤 동일한 볼륨을 마운트하여 다시 builder를 생성한다.

docker buildx rm builder
builder removed

echo '{"Name":"builder","Driver":"docker-container","Nodes":[{"Name":"builder0","Endpoint":"unix:///var/run/docker.sock","Platforms":null,"Flags":null,"DriverOpts":null,"Files":null}],"Dynamic":false}' >  .docker/buildx/instances/builder

docker run -d -v $HOME/volume:/var/lib/buildkit --privileged --name buildx_buildkit_builder0 moby/buildkit:buildx-stable-1
0609d310445aab9309f5b3f3fda9dd0b70fba476cf770a4596f8015378c303ef

이제 동일한 Dockerfile로 빌드를 해서 빌드캐시가 잘 동작하는지 확인한다.

docker buildx build .
WARNING: No output specified with docker-container driver. Build result will only remain in the build cache. To push result image into registry use --push or to load image into docker use --load      
[+] Building 11.0s (9/9) FINISHED
 => [internal] load .dockerignore                                                              0.0s 
 => => transferring context: 2B                                                                0.0s 
 => [internal] load build definition from Dockerfile                                           0.0s 
 => => transferring dockerfile: 361B                                                           0.0s 
 => [internal] load metadata for docker.io/library/python:3.7                                  1.9s 
 => [1/6] FROM docker.io/library/python:3.7@sha256:bf85a74f4ace82f3503c2199aaae10e7a8e370bc7e  0.0s 
 => => resolve docker.io/library/python:3.7@sha256:bf85a74f4ace82f3503c2199aaae10e7a8e370bc7e  0.0s 
 => CACHED [2/6] RUN mkdir /home/ubuntu                                                        0.0s 
 => CACHED [3/6] WORKDIR /home/ubuntu                                                          0.0s 
 => CACHED [4/6] RUN apt-get update && apt-get install -y libgl1-mesa-dev                      0.0s 
 => CACHED [5/6] RUN pip install Tensorflow==2.2                                               0.0s 
 => [6/6] RUN pip install opencv-python==4.5.1.48                                              9.0s