Evidence kitevidence-folder-template/_filled-example/05-restore-test.md
Filled example

Restore test (filled example)

Acme runs a quarterly restore from the production Cloud SQL backups into an isolated environment, measures the outcome, and files the report. This is the artifact most teams lack and most buyers ask…

Illustrative example for a fictional company. See the blank template at ../05-backups-recovery/restore-test.md.

Acme runs a quarterly restore from the production Cloud SQL backups into an isolated environment, measures the outcome, and files the report. This is the artifact most teams lack and most buyers ask for.

Evidence on file

Artifact What it shows
restore-test-2026-06-14.md The completed Q2 test report (below)
restore-runbook.md The procedure followed

Q2 restore test report

Tester: L. Favre · Reviewer: R. Mensah

Item Value
Source backup Automated snapshot 2026-06-13T03:00Z
Source DB prod-main (Cloud SQL Postgres 16)
Target environment restore-test (same region, isolated VPC)
Restore initiated 2026-06-14T14:00Z
Restore completed 2026-06-14T14:41Z
Wall-clock time 41 min
Row count: users 211,904 (expected ~211,900; delta 0.002%)
Row count: events 8,442,118 (expected ~8,442,000; delta 0.001%)
FK integrity PASS
Application boot PASS — read-only smoke test against restored DB
Tear-down Completed 2026-06-14T15:05Z

Outcome: PASS. Restore time 41 min, well within RTO (4 h). No integrity issues.

RPO / RTO answer

RPO: 5 minutes — point-in-time recovery is enabled; worst-case data loss is one WAL interval.

RTO: 4 hours — validated by the quarterly restore test (most recent 2026-06-14, 41 min wall-clock, comfortably under target).

Refresh

Quarterly. Last refreshed 2026-06-14 by L. Favre. Next due 2026-09-14.

Book fit call