Interview preparation

Answer with architecture, evidence and trade-offs.

Weak answers name tools. Strong answers explain controls, failure modes, operational evidence and why the design fits the risk.

01

How would you design a production CI/CD pipeline?

Start with source control, required checks, reproducible builds, tests, security scans, immutable artifacts, environment approvals, deployment strategy, telemetry and rollback.

02

What is the difference between continuous delivery and continuous deployment?

Continuous delivery keeps every validated change releasable; continuous deployment automatically releases every change that passes the controls.

03

How do you manage Terraform state safely?

Use a remote encrypted backend, locking, least privilege, environment separation, versioning, backup and controlled state operations.

04

How do you troubleshoot a failing Kubernetes deployment?

Inspect rollout status, events, pod state, logs, probes, resource limits, configuration, secrets, networking and dependencies before choosing rollback or repair.

05

What makes an alert actionable?

It identifies user impact, has a clear owner, avoids duplicate noise, includes context and links to a tested response procedure.

06

How do you secure a software supply chain?

Protect branches, pin trusted actions, scan dependencies and images, manage secrets externally, generate provenance and SBOMs, sign artifacts and enforce policy.