gigl.orchestration.img_builder#

Attributes#

CUSTOMER_SRC_DOCKERFILE_PATH

logger

Functions#

build_and_push_customer_src_images(context_path, ...)

Package user provided code located at context_path into docker images based on the base images provided.

build_and_push_image(base_image, image_name, ...[, ...])

Builds and pushes a Docker image.

Module Contents#

gigl.orchestration.img_builder.build_and_push_customer_src_images(context_path, export_docker_artifact_registry, base_image_cuda=DOCKER_LATEST_BASE_CUDA_IMAGE_NAME_WITH_TAG, base_image_cpu=DOCKER_LATEST_BASE_CPU_IMAGE_NAME_WITH_TAG, base_image_dataflow=DOCKER_LATEST_BASE_DATAFLOW_IMAGE_NAME_WITH_TAG)[source]#

Package user provided code located at context_path into docker images based on the base images provided. The images are pushed to the export_docker_artifact_registry.

Parameters:

  • context_path (str) – Root directory that will be copied into the docker images.

  • export_docker_artifact_registry (str) – Docker artifact registry to push the images to.

  • base_image_cuda (str) – Base image to use for the CUDA image.

  • base_image_cpu (str) – Base image to use for the CPU image.

  • base_image_dataflow (str) – Base image to use for the Dataflow image.

Returns:

The names of cuda, cpu, and dataflow images.

Return type:

tuple[str, str, str]

gigl.orchestration.img_builder.build_and_push_image(base_image, image_name, dockerfile_path, context_path, multi_arch=False)[source]#

Builds and pushes a Docker image.

Parameters:

  • base_image (Optional[str]) – The base image to use for the build.

  • image_name (str) – The name of the Docker image to build and push.

  • dockerfile_path (str) – The path to the Dockerfile to use for the build.

  • context_path (str) – The path to the context to use for the build.

  • multi_arch (bool) – Whether to build a multi-architecture Docker image. Defaults to False.

Return type:

None

gigl.orchestration.img_builder.CUSTOMER_SRC_DOCKERFILE_PATH[source]#
gigl.orchestration.img_builder.logger[source]#