Skip to main content

Capstone restore and readiness evidence

This required 22–34-hour integrated package strengthens the existing guided labs. Allocation: lessons/predictions 5–7 hours, restore/control experiments 11–19, assessment/repair 6–8. Author estimates require learner timing pilots and are included in the semester workload contract. Reuse the same service and evidence; avoid creating a second portfolio.

Prerequisite bridge​

Bring the preceding module's implementation, risk record and failure observations. Separate assumptions, model results and actual runtime evidence. Review the specific prerequisite named in each lesson before its exercise. Missing runtime tools leave execution evidence pending; a written prediction is not an executed experiment.

Lesson sequence​

  1. Measure recovery from a declared incident boundary.
  2. Tie each security assertion to a tested boundary.
  3. Make the readiness decision survive a game day.

Integrated experiment​

Reuse the original SLO, threat/access review, incident and restore exercises. Add a known synthetic commit sequence for RPO, a fully defined recovery timer, positive/negative tenant tests and a second-operator runbook retest. No real incident messages are sent by this authoring task; learners can rehearse communications locally with clearly fictional recipients.

Complete the assessment. Original guided labs, security controls, deployment/testing and peer requirements remain mandatory. Use synthetic data and a disposable local environment where supported; no cloud purchase is required by this extension.

Selective resources​

Google SRE emergency response; OWASP ASVS; NIST SSDF 1.1.

Runnable restore bridge​

Use restore-evidence.zip with Python 3 standard library before the capstone restore. Predict the restored records, run python3 -B exercise.py, and explain why a valid SQLite file can still contain the wrong application data. The archive performs a real disposable backup/restore and a wrong-data negative control. Its tiny local timing is not a production RTO or a substitute for your own full-service restore. Extend it with a second snapshot and wrong-snapshot selection before completing OR3.

Reference: Python SQLite backup API and SQLite integrity check. Application-record validation remains separate from storage integrity.