AZ400-Q: make deployment retries safe
A deployment can fail after creating some resources but before completing. The release system may retry the job. What property should the deployment automation have?
- A) Idempotency: rerunning with the same desired state converges safely instead of duplicating or corrupting resources. ✓
- B) A random name for every resource on each retry.
- C) An instruction to ignore failed jobs.
- D) A longer pull-request description.
Answer: A. Idempotent deployment makes retries and recovery predictable. Declarative IaC commonly supports this when designed with stable identities and desired state.
Why not the others: Random creation causes drift. Ignoring failures and extra documentation do not make execution safe.
Related
pipeline-reliability · deployment-automation · rollback-strategy