Backup / Restore Readiness Check
Verify that a self-hosted service has a restorable backup before upgrades or destructive maintenance.
Evidence tier describes my hands-on relationship with the content. Official sources describe platform contracts/documentation. Production observations should not automatically be treated as official limits.
Identify state
List databases, named volumes, bind mounts, configuration and external credentials needed to reconstruct the service.
Create a dated backup
Export or snapshot the state through an application-safe method where possible.
Store it outside the live volume
A copy inside the same filesystem is not protection from disk or host failure.
Record versions and config
Capture application/container versions and non-secret configuration required by restore.
Perform a restore test
A backup is an assumption until one representative restore succeeds.
Define retention and rotation
Keep enough history to recover from corruption discovered after the latest backup.