AWS, Azure & GCP · Cloud across providers
How does CI/CD work with cloud infrastructure?
Short answer
CI builds and tests on every change; CD deploys through environments with approvals and automated rollback. Cloud pipelines use OIDC federation for short-lived credentials instead of stored long-lived keys.
A cloud pipeline has two flows: application delivery and infrastructure delivery. Both need the same properties: reproducible builds, environment promotion, approvals where risk justifies them, and a fast rollback path.
Pipeline stages
- Build and unit test on every commit
- Scan dependencies, containers and IaC for vulnerabilities and policy violations
- Plan infrastructure on pull requests and apply after merge
- Deploy progressively with canary or blue-green where possible
- Run smoke tests and roll back automatically on failure
- Authenticate with OIDC federation rather than stored cloud keys
permissions:
id-token: write
contents: read
steps:
- uses: aws-actions/configure-aws-credentials@v4
with:
role-to-assume: arn:aws:iam::123456789012:role/deploy
aws-region: ap-south-1Separate build from deploy and keep environment credentials scoped per environment. A pipeline that can deploy to production on every branch is a security and reliability risk. Immutable image tags and versioned artifacts make rollbacks deterministic.
Follow-ups you should expect
Interviewers use these to check whether the first answer was understanding or memorisation.
- 1Why use OIDC instead of storing cloud access keys in CI?
- 2How do you roll back a bad infrastructure change?
- 3What is the difference between canary and blue-green deployment?
- 4How do you keep environment credentials scoped?
Say it in the interview
CI builds and tests every change, and CD promotes artifacts through environments with approvals and automated rollback. I authenticate pipelines with OIDC so they exchange a token for short-lived cloud credentials instead of storing keys. I also plan infrastructure on pull requests, apply after merge, and verify health after deploy rather than assuming a green pipeline means success.
Use this as a model, not a script. Saying it in your own words is what makes it survive follow-up pressure.
Related questions
Reading is step one
Now say it out loud.
Understanding an answer and delivering it under pressure are different skills. Rehearse this topic in a live AWS, Azure & GCP mock interview and get scored feedback on depth, clarity and confidence.