When the primary artifact repository becomes unavailable, a contingency plan relies on redundancy and caching so deployments aren't blocked by a single point of failure.
Key Points: • Maintain a secondary, synchronized artifact repository as a fallback (many repository managers support built-in mirroring/replication). • Keep critical artifacts cached locally on build agents or in a distributed cache to survive short outages. • Pin CI/CD pipelines to retry against the fallback repository automatically if the primary times out. • Regularly test failover to the secondary repository so it's a proven path, not just a theoretical one. • Document a manual override process for deploying a known-good, previously pulled artifact if both repositories are down.
Example: If the primary Nexus instance goes down mid-release, the deployment pipeline automatically retries against a replicated secondary Nexus in another region, so the release proceeds with only a brief delay.
Interview Tip: A concise interview answer is:
"I'd have a synchronized secondary artifact repository as a fallback, plus local or distributed caching of recently used artifacts so short outages don't block deployment. The pipeline should be able to fail over automatically, and that failover path should be tested periodically, not just assumed to work."