Workbench Base Docker Build and Push
Notes and information on how to do the Docker Builds and Push to AWS ECR.
Update Workbench Version
vi Dockerfile
# Install latest Workbench
RUN pip install --no-cache-dir 'workbench[ml-tool,chem]'==0.7.0
Build the Docker Image
Note: For a client specific config file you'll need to copy it locally so that it's within Dockers 'build context'. If you're building the 'vanilla' open source Docker image, then you can use the open_source_config.json
that's in the directory already.
docker build --build-arg WORKBENCH_CONFIG=open_source_config.json -t \
workbench_base:v0_7_0_amd64 --platform linux/amd64 .
Test the Image Locally
You have a docker_local_base
alias in your ~/.zshrc
:)
Login to ECR
aws ecr-public get-login-password --region us-east-1 --profile \
scp_sandbox_admin | docker login --username AWS \
--password-stdin public.ecr.aws
Tag/Push the Image to AWS ECR
Update the 'latest' tag
docker tag public.ecr.aws/m6i5k1r2/workbench_base:v0_7_0_amd64 \
public.ecr.aws/m6i5k1r2/workbench_base:latest
Update the 'stable' tag
This is obviously only when you want to mark a version as stable. Meaning that it seems to 'be good and stable (ish)' :)
docker tag public.ecr.aws/m6i5k1r2/workbench_base:v0_7_0_amd64 \
public.ecr.aws/m6i5k1r2/workbench_base:stable
Test the ECR Image
You have a docker_ecr_base
alias in your ~/.zshrc
:)