Introduction

A release is deployed to production by running a deployment script. The input for that script is the list of git commit id of each component that is going to be released.

That git commit compositions of a release must be recorded, so we know which component version combination will result a specific QA test result.

In order to improve efficency and to avoid mistakes while:

• tracking release component combinations
• giving input to deployment script
• mapping QA test result with specific released component combinations

We need to:

• integrate all created documents to the same database.
• use that database to automate the generation of deployment script input.
• use that database to automate the generation of the form to be used by QA to enter test results.

The Sequence Diagram

Zero-Ambiguity Release Verification in CI/CD

Exposing the deployed Git commit SHA and build timestamp via an authenticated health/version endpoint allows automated post-deployment smoke tests to confirm when a rolling Cloud Run or Kubernetes release has completely replaced older pods.