Skip to main content

Cloud failure-domain evidence

This required 20–30-hour integrated package extends guided Labs 2–4: lessons/predictions 5–7 hours, model and topology/inventory evidence 9–15, assessment/repair 6–8. Reuse the original architecture and cost decision; these author estimates will be reconciled at semester integration and require learner timing pilots.

Prerequisite bridge​

Review Semester 05 routes and request paths, Semester 07 dependency graphs, and Semester 08 usable capacity and missing evidence. Distinguish request rate from storage volume and service dependency from network reachability. Complete the original identity boundary lab; this model does not evaluate IAM.

Sequence​

  1. Dependency closure.
  2. Surviving capacity.
  3. Inventory and cost boundaries.
  4. Download failure-domain-contracts.zip, implement the Python standard-library starter and run python3 -B check.py starter before reference calibration.
  5. Complete CF1–CF6 using original guided labs as well as the model.

The archive checks 6,144 finite scenarios and reversed insertion order, plus invalid inputs. It assumes instantaneous traffic redistribution, additive per-service capacity and an acyclic graph of mandatory dependencies. It does not emulate cloud services, network routing, IAM, data replication, provisioning or failure probabilities. No cloud account, credentials or purchase is needed.

For the original build-or-model lane, attach manually traced routes, allowed/denied policy cases and a synthetic inventory with an incomplete-query case. Clearly label modeled results. If using actual services later, retain real permission, failure and teardown evidence; a model pass never proves deployed behavior.