Git, pipelines, Terraform, Ansible
Practise these as flashcards. One or two lines each; the ones you miss come back first.
Start freeContinuous Integration builds and tests code automatically on every commit so problems show up early. Continuous Delivery promotes a tested build through environments automatically, with approval gates where the business wants them. From infra, I make sure the runners, targets, credentials and network paths are in place and monitored.
The packaged output of a build, like a JAR, a zip or a container image, stored with a version in a repository so the exact same thing goes to every environment.
A request to merge a branch into main, with code review and automated checks first. For infra code it works like a change request.
Rolling updates a few servers at a time. Blue/green runs two full environments and flips the load balancer. Canary sends a small slice of traffic to the new version first. All of them rely on the LB and health checks.
Can the build agent reach the target on that port (SSH 22, WinRM 5985/5986)? Usually a missing firewall rule from the agent subnet, the service not running on the target, or wrong credentials. Test from the agent with nc -zv.
Terraform provisions infrastructure (VMs, networks, load balancers) and tracks it in state. Ansible configures what runs on the machines. They're often used together.
State maps your code to real resources. Kept remote (Azure Storage, S3) with locking, the team shares one source of truth and two people can't apply at the same time and corrupt it.
Running it once or ten times leaves the system in the same state. Ansible modules check before changing; a plain shell script usually doesn't.
Builds are repeatable and reviewable, changes are version-controlled, environments stay identical, and rollback is reverting a commit. No hand-built snowflake servers.
Never in the repo. Store them in a vault (Azure Key Vault, HashiCorp Vault) or the CI tool's credential store, inject them at runtime, mask them in logs, and prefer short-lived identities like OIDC or managed identities over long-lived keys.