Deploying the app on IBM Cloud

In this guide, you’ll learn how to:

Push the previously built container image to IBM Container Registry (ICR).
Deploy the container to IBM Code Engine, a fully managed, serverless platform for running containers.

This workflow is ideal for hosting custom services like a secured API backend, built and maintained locally but deployed in the cloud.

Prerequisites

Python container image built (previous step)
IBM Cloud CLI installed
Logged in with ibmcloud login --sso

Install required plugins

Install the Code Engine and Container Registry plugins:

ibmcloud plugin install code-engine container-registry

Set environment variables

These variables make the rest of the commands easier to reuse. Choose a unique IMAGE_NAME

export RESOURCE_GROUP=rotkreuz-dibm
export CR_NAMESPACE=dibm
export IMAGE_NAME=backend-${USER}
export IMAGE_TAG=1.0
export PROJECT_NAME=dibm

Note: These variables are only available in the current session.

Setting environment variables on Windows

Check the Windows command sections in the pre-requisites chapter for instructions on how to execute this on Windows. The USER environment variable is automatically set to your username on macOS and Linux. Wherever you see $USER, use USERNAME on Windows. Below an example for Windows Powershell:

$env:IMAGE_NAME = "backend-$env:USERNAME"

Select IBM Cloud context

Select the correct resource group:

ibmcloud target -g ${RESOURCE_GROUP}

Select your Code Engine project:

ibmcloud ce project select --name ${PROJECT_NAME}

Publish image to IBM Container Registry

Set region and log in:

ibmcloud cr region-set eu-central
ibmcloud cr login --client podman # or --client docker

Add namespace (skip if it already exists):

ibmcloud cr namespace-add ${CR_NAMESPACE}

A manifest groups multiple platform-specific container images under one tag, allowing cross-architecture support (e.g., amd64, arm64). We re-tag the image with the IBM Container Registry URL so it can be correctly identified and pushed to the registry.

podman tag backend:1.0 de.icr.io/${CR_NAMESPACE}/${IMAGE_NAME}:${IMAGE_TAG}
podman manifest push de.icr.io/${CR_NAMESPACE}/${IMAGE_NAME}:${IMAGE_TAG}

(Optional) Set image retention policy to keep only the last 2 versions:

ibmcloud cr retention-policy-set --images 3 ${CR_NAMESPACE}

Create API key and registry access for Code Engine

Create an IBM Cloud API key:

Save the generated API key, as it’s only visible during creation time!

ibmcloud iam api-key-create backend-deploy-key-${USER} -d "API Key to deploy a Python backend on IBM Code Engine"

Copy and export the API key:

export API_KEY="myGeneratedAPIKey"

Create a registry access secret for Code Engine:

ibmcloud ce registry create --name ibm-container-registry-${USER} --server de.icr.io --username iamapikey --password ${API_KEY}

Create the API credential secret

ibmcloud ce secret create --name api-credentials-${USER} --from-literal UN=myUsername --from-literal PW=myPassword

Deploy the Python app to IBM Code Engine

Run the following to create and deploy the application:

ibmcloud ce app create --name backend-${USER} \
  --image de.icr.io/${CR_NAMESPACE}/${IMAGE_NAME}:${IMAGE_TAG} \
  --registry-secret ibm-container-registry-${USER} \
  --env-from-secret api-credentials-${USER} \
  --port 8000 \
  --min-scale 1 \
  --max-scale 1 \
  --cpu 0.25 \
  --memory 0.5G

This will start your Python backend on IBM Code Engine.

Container to Cloud

Workshop documentation to build a local container image and deploy it ti IBM Code Engine