▶GitHub Actions vs Jenkins in 2026 — which should I learn?
GitHub Actions: cloud-native, minimal setup, free for public repos, tightly integrated with GitHub workflows, best for startups/small teams. Jenkins: self-hosted, maximum flexibility, massive plugin ecosystem, steeper learning curve but dominates enterprises (Fortune 500). New projects: GitHub Actions. Existing enterprise: Jenkins. Both → +$40k salary lift.
▶Blue-green vs canary deployments — when do I use each?
Blue-green: instant zero-downtime cutover (run two full environments, switch traffic). Requires 2x infra cost, safest for critical systems. Canary: route 5-10% traffic to new version, monitor metrics, gradually shift. Cheaper but slower feedback. For APIs: canary (10min rollout). For UX-breaking changes: blue-green (instant rollback). Most teams use both.
▶How do I manage secrets in CI/CD pipelines safely?
Never commit .env files. Use: (1) GitHub Secrets / GitLab CI Variables (encrypted at rest, masked in logs), (2) HashiCorp Vault (dynamic creds, audit trail), (3) AWS Secrets Manager / Parameter Store (cloud-native, IAM-based access). Rotate secrets every 90 days. Audit access logs. HashiCorp Vault is overkill for < 50 engineers.
▶What's the difference between build caching strategies?
Layer caching: Docker caches each Dockerfile line; abuse it by putting dependencies early. Artifact cache: store compiled binaries/node_modules between jobs (faster than reinstall). Git shallow clone + dependency lock files = huge speedups. Cache key strategy: `key: ${{ runner.os }}-npm-${{ hashFiles('**/package-lock.json') }}`. Smart cache = 5-10x faster pipelines.
▶When should I outsource CI/CD vs build in-house?
Outsource (GitHub Actions, CircleCI, GitLab CI) if: < 50 engineers, < 10 parallel builds/day, simple workflows. Build in-house (Jenkins on k8s) if: > 200 engineers, complex build matrix, strict compliance/air-gap. Hybrid: GitHub Actions for unit tests, Jenkins for deploy approval gating.
▶What's GitOps and why do I care?
GitOps: Git is the single source of truth for all infra + app config. Declarative (git push = deploy, no manual steps). Tools: ArgoCD, Flux. Benefits: audit trail, rollback via git revert, reproducible deploys. Downside: steep learning curve, requires deep k8s knowledge. For startups: add GitOps in Phase 2. For enterprises: GitOps is mandatory for compliance.
▶How long does a CI/CD pipeline take to master?
L1 (GitHub Actions/GitLab CI basics): 6-8 weeks, 2-3 hrs/week. L2 (IaC + multi-env): 3-4 months additional. L3 (GitOps + SRE): 6-12 months. Total investment 8-12 months for full mastery. ROI: $25-50k salary bump, 10-20% delivery speed improvement, 80% fewer deploy bugs.