Disaster Recovery & Manual Rescues

🎯 Overview

In the event of a backup failure or system crash, this guide provides the procedures for manually intervening and rescuing the backup state.

🆘 Manual Rescue Workflow

If a backup job is interrupted (e.g., due to a network timeout or system reboot), the Restic repository may remain "locked."

1. Identify the Problem

Check the logs at /var/log/restic-backup.log (daily) or /var/log/monthly-backup.log (monthly). Common issues include:

• repository is already locked: Indicates a previous run crashed.
• rclone: Token expired: Indicates GDrive authentication failure.

2. The Rescue Procedure

Use the standardized workflow defined in the project:

# Unlock the repository
restic -r rclone:gdrive:backups/vps-monthly unlock

# Resume the backup (Restic will deduplicate data)
bash /usr/local/bin/monthly-full-backup.sh

📂 Data Restoration

To restore data to a new system or recover lost files:

List Snapshots

restic -r rclone:gdrive:backups/vps-monthly snapshots

Restore Full Snapshot

restic -r rclone:gdrive:backups/vps-monthly restore latest --target /path/to/restore

Restore Specific Folder

restic -r rclone:gdrive:backups/vps-monthly restore latest --target /tmp/recovery --include /home/christian/data

🧪 Verification

Always run backup-health-check.sh on the VPS after a manual rescue to ensure the repository remains consistent.