Manual deployment releases software through commands executed directly by a person. It is the practical baseline for automation: first run the deployment locally, then capture the working steps in scripts that can be tested before production use. section 8.9.3
A workable manual deployment requires:
- A deployment target, such as a Kubernetes cluster running in the cloud, because cloud automation cannot reach a local Kubernetes instance. section 8.9.3
- A container registry for storing the image that will be deployed. section 8.9.3
- A locally tested sequence for building, publishing, and deploying the microservice. section 8.9.3
From commands to repeatable deployment
Set registry and version variables, build a production container image, authenticate to the registry, push the image, and apply the Kubernetes configuration. Once these commands work locally, place them in shell scripts so later changes can be tested before running them through GitHub Actions. section 8.9.3 section 8.6
Browse the image-building command that forms the first step of the scripted deployment sequence. section 8.9.3
