Skip to content

fix: Rename buildspec files. #211

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Jun 10, 2020
Merged

fix: Rename buildspec files. #211

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
1d273ee
infra: Rename buildspec.yml -> buildspec-container-pr.yml.
nadiaya Jun 9, 2020
3c5302b
infra: Rename buildspec-toolkit.yml -> buildspec.yml.
nadiaya Jun 9, 2020
f13ff09
Move lib file from out of docker folder.
nadiaya Jun 10, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 13 additions & 0 deletions buildspec-container-pr.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
version: 0.2

phases:
pre_build:
commands:
- PR_NUM=$(echo $CODEBUILD_SOURCE_VERSION | grep -o '[0-9]\+')
- echo 'Pull request number:' $PR_NUM '. No value means this build is not from pull request.'

build:
commands:

- error_cmd="echo 'In order to make changes to the docker files, please, use https://github.com/aws/deep-learning-containers repository.' && exit 1"
- execute-command-if-has-matching-changes "$error_cmd" "docker/"
94 changes: 0 additions & 94 deletions buildspec-toolkit.yml
View file
Open in desktop

This file was deleted.

85 changes: 83 additions & 2 deletions buildspec.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,13 +1,94 @@
version: 0.2

env:
variables:
FRAMEWORK_VERSION: '1.4.0'
CPU_INSTANCE_TYPE: 'ml.c4.xlarge'
GPU_INSTANCE_TYPE: 'ml.p2.8xlarge'
ECR_REPO: 'sagemaker-test'
GITHUB_REPO: 'sagemaker-pytorch-container'
DLC_ACCOUNT: '763104351884'
SETUP_FILE: 'setup_cmds.sh'
SETUP_CMDS: '#!/bin/bash\npip install --upgrade pip\npip install -U -e .\npip install -U -e .[test]'

phases:
pre_build:
commands:
- start-dockerd
- ACCOUNT=$(aws --region $AWS_DEFAULT_REGION sts --endpoint-url https://sts.$AWS_DEFAULT_REGION.amazonaws.com get-caller-identity --query 'Account' --output text)
- PREPROD_IMAGE="$ACCOUNT.dkr.ecr.$AWS_DEFAULT_REGION.amazonaws.com/$ECR_REPO"
- PR_NUM=$(echo $CODEBUILD_SOURCE_VERSION | grep -o '[0-9]\+')
- BUILD_ID="$(echo $CODEBUILD_BUILD_ID | sed -e 's/:/-/g')"
- echo 'Pull request number:' $PR_NUM '. No value means this build is not from pull request.'

build:
commands:
- TOX_PARALLEL_NO_SPINNER=1
- PY_COLORS=0

# install
- pip3 install -U -e .[test]

# run linters
- tox -e flake8,twine

# run unit tests
- tox -e py27,py36,py37 test/unit

# define tags
- GENERIC_TAG="$FRAMEWORK_VERSION-pytorch-$BUILD_ID"
- DLC_CPU_TAG="$FRAMEWORK_VERSION-dlc-cpu-$BUILD_ID"
- DLC_GPU_TAG="$FRAMEWORK_VERSION-dlc-gpu-$BUILD_ID"

# run local CPU integration tests (build and push the image to ECR repo)
- test_cmd="pytest test/integration/local --build-image --push-image --dockerfile-type pytorch --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor cpu --tag $GENERIC_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"
- test_cmd="pytest test/integration/local --build-image --push-image --dockerfile-type dlc.cpu --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor cpu --tag $DLC_CPU_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"

# launch remote GPU instance
- prefix='ml.'
- instance_type=${GPU_INSTANCE_TYPE#"$prefix"}
- create-key-pair
- launch-ec2-instance --instance-type $instance_type --ami-name dlami-ubuntu-latest

# build DLC GPU image because the base DLC image is too big and takes too long to build as part of the test
- python3 setup.py sdist
- build_dir="test/container/$FRAMEWORK_VERSION"
- $(aws ecr get-login --registry-ids $DLC_ACCOUNT --no-include-email --region $AWS_DEFAULT_REGION)
- docker build -f "$build_dir/Dockerfile.dlc.gpu" -t $PREPROD_IMAGE:$DLC_GPU_TAG --build-arg region=$AWS_DEFAULT_REGION .
# push DLC GPU image to ECR
- $(aws ecr get-login --registry-ids $ACCOUNT --no-include-email --region $AWS_DEFAULT_REGION)
- docker push $PREPROD_IMAGE:$DLC_GPU_TAG

# run GPU local integration tests
- printf "$SETUP_CMDS" > $SETUP_FILE
# no reason to rebuild the image again since it was already built and pushed to ECR during CPU tests
- generic_cmd="pytest test/integration/local --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor gpu --tag $GENERIC_TAG"
- test_cmd="remote-test --github-repo $GITHUB_REPO --test-cmd \"$generic_cmd\" --setup-file $SETUP_FILE --pr-number \"$PR_NUM\""
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"
- dlc_cmd="pytest test/integration/local --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor gpu --tag $DLC_GPU_TAG"
- test_cmd="remote-test --github-repo $GITHUB_REPO --test-cmd \"$dlc_cmd\" --setup-file $SETUP_FILE --pr-number \"$PR_NUM\" --skip-setup"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"

# run CPU sagemaker integration tests
- test_cmd="pytest -n 10 test/integration/sagemaker --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor cpu --instance-type $CPU_INSTANCE_TYPE --tag $GENERIC_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"
- test_cmd="pytest -n 10 test/integration/sagemaker --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor cpu --instance-type $CPU_INSTANCE_TYPE --tag $DLC_CPU_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"

# run GPU sagemaker integration tests
- test_cmd="pytest -n 10 test/integration/sagemaker --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor gpu --instance-type $GPU_INSTANCE_TYPE --tag $GENERIC_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"
- test_cmd="pytest -n 10 test/integration/sagemaker --region $AWS_DEFAULT_REGION --docker-base-name $ECR_REPO --aws-id $ACCOUNT --framework-version $FRAMEWORK_VERSION --processor gpu --instance-type $GPU_INSTANCE_TYPE --tag $DLC_GPU_TAG"
- execute-command-if-has-matching-changes "$test_cmd" "test/" "src/*.py" "setup.py" "setup.cfg" "buildspec-toolkit.yml" "docker/build_artifacts/*"

finally:
# shut down remote GPU instance
- cleanup-gpu-instances
- cleanup-key-pairs

- error_cmd="echo 'In order to make changes to the docker files, please, use https://github.com/aws/deep-learning-containers repository.' && exit 1"
- execute-command-if-has-matching-changes "$error_cmd" "docker/"
# remove ECR image
- aws ecr batch-delete-image --repository-name $ECR_REPO --region $AWS_DEFAULT_REGION --image-ids imageTag=$GENERIC_TAG
- aws ecr batch-delete-image --repository-name $ECR_REPO --region $AWS_DEFAULT_REGION --image-ids imageTag=$DLC_CPU_TAG
- aws ecr batch-delete-image --repository-name $ECR_REPO --region $AWS_DEFAULT_REGION --image-ids imageTag=$DLC_GPU_TAG