Disaster Recovery Summary

Disaster Recovery Summary

Version: 1.0 · Owner: Pocket PMO Ltd · Approval status: Approved · Review date: annually

Purpose

This summary explains how Pocket PMO would recover from a major service-impacting event, including data loss, hosting provider incident, or credential compromise.

Scenarios covered

1. Regional platform-provider outage.
2. Database corruption or accidental destructive change.
3. Compromise of a highly-privileged credential.
4. Loss of key third-party integration (Stripe, Resend, AI Gateway).

Recovery approach

• 1. Regional outage. Wait for provider restoration; the client PWA remains partly usable through offline caching. If a multi-region option becomes available, we will evaluate and publish adoption.
• 2. Database corruption / destructive change. Restore from Lovable Cloud's managed daily backups (approximately 14 days of retention) to a point immediately prior to the change.
• 3. Credential compromise. Rotate secrets via the platform's secrets tools (LOVABLE_API_KEY via the rotate tool; other secrets via the vault). Force sign-out of active sessions; review audit_logs for scope of exposure. Notify customers per the Incident Response Policy.
• 4. Third-party outage. Fail feature into degraded mode (queue emails for later delivery; disable AI feature gracefully; disable card payments). Publish status.

Recovery objectives

• RTO — 24 hours.
• RPO — 24 hours.
• Backup retention — approximately 14 days of daily backups on Lovable Cloud, browsable and restorable.

Testing

An annual restore test is recommended, using the managed platform's backup restore capability, and documented on this page.

Document history