A disaster is any failure that exceeds normal resilience.

It is not limited to fires and floods. It includes deletion, corruption, control-plane loss, ransomware, and region-wide outages.

“Most real disasters start as a small human or software mistake with a large blast radius.”

Human error

Accidental delete

Wrong target, broad script, or destructive change in production.

Software failure

Silent corruption

Data drift or logic bugs that replicate bad state everywhere.

Platform failure

Cluster unusable

Kubernetes, storage, DNS, or IAM controls no longer recover the app.

Security event

Ransomware or credential theft

Attackers target production and backup paths together.

Provider event

Region degradation

A whole cloud region or service plane becomes unreliable.

Physical event

Power, flood, fire

The classic case is real, just less common than operational mistakes.