# Backup & Recovery

## Backup Strategy

| Data Type | Frequency | Retention |
|-----------|-----------|-----------|
| Database (full) | Daily | 30 days |
| Database (incremental) | Hourly | 7 days |
| Uploaded documents | Daily | 1 year |
| Configuration | On change | 90 days |
| Blockchain state | Daily | 90 days |

## Database Backup

### Automated Backups

```bash
# Daily full backup (runs at 02:00 UTC)
pg_dump -Fc tlas_prod > /backups/tlas_$(date +%Y%m%d).dump

# Hourly WAL archiving
archive_command = 'cp %p /backups/wal/%f'
```

### Manual Backup

```bash
# Create backup
docker exec postgres pg_dump -U tlas -Fc tlas_prod > backup.dump

# Verify backup
pg_restore --list backup.dump
```

## Recovery Procedures

### Database Recovery

```bash
# Stop application
docker-compose stop api

# Restore database
pg_restore -d tlas_prod -c backup.dump

# Verify data
psql -d tlas_prod -c "SELECT COUNT(*) FROM applications;"

# Restart application
docker-compose start api
```

### Point-in-Time Recovery

```bash
# Restore to specific timestamp
recovery_target_time = '2026-02-09 10:00:00'
```

## Disaster Recovery

### RTO/RPO Targets

| Metric | Target |
|--------|--------|
| Recovery Time Objective (RTO) | 4 hours |
| Recovery Point Objective (RPO) | 1 hour |

### DR Procedure

1. Identify failure scope
2. Activate DR environment
3. Restore from latest backup
4. Verify data integrity
5. Update DNS to DR site
6. Notify stakeholders

## Backup Verification

Monthly backup testing:
- Restore to test environment
- Run integrity checks
- Verify application functionality
- Document results