Backup and Data Recovery
Torneyo's backup policy: frequency, retention, scope, quarterly restore testing, RPO/RTO commitments, and self-service full data export.
This document formalizes the backup and data recovery policy of the Torneyo platform (torneyo.com and app.torneyo.com): what is backed up, how often, how long it is retained, how restoration is tested, and which recovery objectives are committed to. It can be attached to proposals, contracts, and public procurement processes as a formal commitment, together with the Service Level Agreement, Privacy and Data Protection (LGPD), the Technical Specification, and Infrastructure and Data Location.
Backup scope
| Asset | Where it resides | Protection mechanism |
|---|---|---|
| Database (records, competitions, results, billing, audit trail) | Managed PostgreSQL database (Amazon RDS), São Paulo region — see Infrastructure and Data Location | Automated daily backup (snapshot) + point-in-time recovery |
| Files (photos, scanned documents, match sheets, attachments) | Object storage service (Cloudflare R2) | Multiple redundancy maintained by the provider, with enterprise-grade durability; file references included in the database backup |
| Infrastructure configuration | Versioned code repository | Infrastructure described as code, rebuildable from the repository |
Frequency and method
- Automated daily database backup: a full snapshot is taken automatically by the managed service, every day, with no manual intervention.
- Point-in-time recovery (PITR): in addition to the daily snapshot, the managed service continuously retains transaction logs, allowing the database to be restored to any moment within the retention period — not only to the last snapshot.
- On-demand snapshots: before maintenance with potential structural impact on data (significant schema migrations), an additional manual snapshot is taken.
Retention
- Automated database backups are retained for at least 7 days, during which point-in-time recovery is available.
- Contracts requiring longer retention (30 days or more, or long-term copies) can include it as a project-specific condition, formalized in a proposal.
- Data deleted at the request of the data subject or the controlling organization leaves the backups as rotation cycles expire, as described in the privacy policy.
Recovery objectives (RPO and RTO)
| Objective | Commitment | Note |
|---|---|---|
| RPO (maximum data loss) | 24 hours | Formal commitment based on the daily cycle; in practice, point-in-time recovery typically allows restoring the database with only a few minutes of loss |
| RTO (maximum time to restore) | 4 business hours | Counted from the decision to restore, for the full database restoration scenario |
The values above are honest, verifiable commitments for the platform’s current scale. Scenarios requiring more aggressive RPO/RTO (real-time replica, multi-zone high availability) can be addressed as project-specific conditions — see the infrastructure transparency note in the SLA.
Periodic restore testing
There is no reliable backup without tested restoration. Torneyo commits to performing, at least once per quarter, a database restore test in an isolated environment, recording:
- date and time of the test;
- backup used (snapshot or point-in-time);
- result (intact/with caveats) and total restoration time.
Test records can be presented to the contracting party upon request through the formal support channel (contato@torneyo.com).
Self-service logical backup (full data export)
Independently of infrastructure backups, the contracting organization can generate its own logical backup at any time, without depending on Torneyo: the organization area offers a full data export in open formats (CSV/JSON/XML, with a data dictionary), available under Organization → Data export in the app. This guarantees portability and eliminates vendor lock-in, as detailed in the privacy policy and in Organizations, members and permissions.
Responsibility
The execution of this policy is the responsibility of the Torneyo technical team, coordinated by the data protection officer named in the privacy policy. Incidents involving data loss follow the incident communication flow described there (notification to the controlling organization within 24 hours).
Revision history
| Version | Date | Changes |
|---|---|---|
| 2026.08.19 | 2026-08-19 | Initial version of the document. |