SAN change and rollback

Outcome

Execute a bounded, observable change with explicit validation and a rehearsed route back to the known-good state.

Change contract

Document purpose, exact objects, dependency map, pre-check evidence, risk, window, approver, executor, observer, success gates, abort thresholds, rollback steps, and communications.

Procedure

1. Capture configuration and health state; verify no conflicting incident or change is active.
2. Validate every identity and confirm the intended active configuration.
3. Simulate or peer-review the exact command sequence and rollback.
4. Apply the smallest atomic change to one fault domain first where architecture permits.
5. Validate effective policy, logins, paths, application I/O, and new error-counter deltas.
6. Continue only when success gates pass; otherwise stop and roll back.
7. Capture post-change state and monitor through the declared observation window.

Rollback triggers

Unexpected visibility, path loss beyond the intended domain, new errors, application impact, configuration mismatch, monitoring blindness, or elapsed time exceeding the safe window.

Production boundary

Deletion and activation commands are intentionally simulated in StorageCraft. Real changes require platform-specific backups, approvals, syntax, and operational authority.

Next step

Practice create, activate, validate, delete, and restore workflows in the Storage Network Academy.