What is IaC and its CI/CD benefits over manual provisioning?

WHAT THIS TESTS: This question checks whether you see Infrastructure as Code as a foundational DevOps practice rather than just automation trivia. The interviewer wants to know if you understand that IaC uses versioned, descriptive models to define networks, virtual machines, load balancers, and connection topologies, and that it is a core component of continuous delivery. They are also listening for your ability to contrast this with manual provisioning and to articulate specific pipeline benefits like repeatability and drift prevention.

A GOOD ANSWER COVERS: First, define IaC as the practice of managing infrastructure through machine-readable definition files rather than manual configuration. Second, emphasize the declarative approach where files specify desired state and components, not step-by-step installation procedures. Third, explain idempotence as the principle that a deployment command always produces the same environment regardless of the starting state, either by reconfiguring the existing target or recreating it fresh. Fourth, connect these concepts to CI/CD by noting that pipelines can version infrastructure alongside application code, spin up production-like test environments on demand, and tear them down automatically, which prevents the snowflake environment problem and eliminates configuration drift between stages.

COMMON WRONG ANSWERS: A weak answer treats IaC as simply writing bash or PowerShell scripts to install software. Another red flag is focusing only on speed without mentioning consistency or idempotency. Candidates sometimes confuse imperative and declarative approaches, describing IaC as a sequence of commands rather than a desired-state model. Avoid suggesting that IaC is only about cloud resources; it also covers on-premises and hybrid topologies. Finally, do not claim that IaC removes the need for testing; in fact, infrastructure code itself should be validated and tested.

LIKELY FOLLOW-UPS: An interviewer may ask how you would test infrastructure code before deployment, or which tools you prefer for declarative IaC such as ARM templates, Bicep, or Terraform. They might probe how you handle secrets and sensitive values in versioned infrastructure definitions, or how you manage state files in a team setting. Another common thread is how you roll back infrastructure changes when a deployment fails, or how you enforce compliance and detect misconfiguration automatically.

ONE CONCRETE EXAMPLE: Imagine a team that manually configured three deployment environments for a web application. Over time, the staging server received a patch that production did not, causing a release to fail only in production. By moving to IaC using JSON-based ARM templates stored in the same repository as the application code, the team ensures that every pull request triggers a pipeline validating both the app and its required Azure infrastructure. When developers need a temporary integration environment, the pipeline provisions it from the same template and destroys it after tests pass, guaranteeing that what ran in testing matches production exactly.

Source: learn.microsoft.com