[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"glossary-disaster-recovery::en":3,"gloss-cluster-disaster-recovery::en":26,"gloss-next-disaster-recovery::en":9},{"slug":4,"category":5,"name":6,"definition":7,"meta_desc":8,"faq":9,"schema_markup":9,"related":10},"disaster-recovery","cloud","Disaster Recovery","Disaster recovery is the plan and the tested capability for bringing a service back after a failure large enough that normal redundancy does not cover it: a region outage, a corrupted primary database, a destructive deploy, or a ransomware event. It is defined by two numbers agreed with the business before anything is built. The recovery time objective is how long the service may be down. The recovery point objective is how much data you can afford to lose, measured backwards from the moment of failure. Everything else — replication topology, backup frequency, whether you run warm standby infrastructure — follows from those two targets, and quoting them without a cost estimate is how plans become fiction. Backups are not disaster recovery on their own. A backup that has never been restored is an assumption, not a capability, and the failures found in a first real restore are consistently the same: a missing credential, an unrecorded schema migration, a dependency on a service that is also down, and a restore that takes far longer than anyone estimated because nobody timed it. Backups also need to be protected from the same event they insure against, which is why immutable and separately-credentialed copies matter for the ransomware case. The only meaningful test is a rehearsal: restore into a clean environment on a schedule, measure the elapsed time against the stated objective, and treat the gap as a defect to fix rather than a number to explain.","Disaster recovery is defined by recovery time and recovery point objectives — and why an untested backup is an assumption rather than a real capability.",null,[11,14,17,20,23],{"slug":12,"name":13},"availability-zone","Availability Zone (AZ)",{"slug":15,"name":16},"backfill","Backfill",{"slug":18,"name":19},"incident-management","Incident Management",{"slug":21,"name":22},"multi-region","Multi-Region",{"slug":24,"name":25},"uptime","Uptime",[27,31,32,35,39,42,45,48,51,54,55,58],{"slug":28,"category":5,"name":29,"updated_at":30},"autoscaling","Autoscaling","2026-08-24T02:46:37+00:00",{"slug":12,"category":5,"name":13,"updated_at":30},{"slug":33,"category":5,"name":34,"updated_at":30},"block-storage","Block Storage",{"slug":36,"category":5,"name":37,"updated_at":38},"edge-ai","Edge AI","2026-08-24T02:46:38+00:00",{"slug":40,"category":5,"name":41,"updated_at":30},"egress-fees","Egress Fees (Data Transfer Out)",{"slug":43,"category":5,"name":44,"updated_at":30},"finops","FinOps (Cloud Financial Operations)",{"slug":46,"category":5,"name":47,"updated_at":38},"immutable-infrastructure","Immutable Infrastructure",{"slug":49,"category":5,"name":50,"updated_at":38},"infrastructure-drift","Infrastructure Drift",{"slug":52,"category":5,"name":53,"updated_at":30},"managed-kubernetes","Managed Kubernetes",{"slug":21,"category":5,"name":22,"updated_at":30},{"slug":56,"category":5,"name":57,"updated_at":38},"noisy-neighbor","Noisy Neighbor",{"slug":59,"category":5,"name":60,"updated_at":30},"platform-as-a-service","Platform as a Service (PaaS)"]