كوارث تحدث فعلًا

في هذا الدرس

list the realistic disasters a subsystem must survive.

الموضوعات المخطط لها

• Common causes are human and software: a bad migration, a deleted volume, a leaked or expired credential, a corrupt deploy, a ransomware or account lockout.
• Infrastructure disasters (zone or region loss) are real but rarer than operator error.
• High availability (surviving a node) is not disaster recovery (surviving the loss of data or a whole site); replication copies mistakes too.
• Scope the subsystem: data stores, event streams, secrets, DNS, configuration, and the people and access needed.