Who This Is For
Intermediate to advanced.
Think of it like a Sunday-night walkthrough of the whole house, lights off, doors locked, fridge stocked, before the week starts. Except a script does the walking for you and just tells you what needs attention.
What You'll Build
A weekly report identifying at least one service, one backup status, one disk-health status, and one action item.
Prerequisites
- NetBox, Restic, and Uptime Kuma already running
Why Visibility Matters
Individually, monitoring, backups, and documentation each answer a narrow question. Combined into one regular report, they answer the question that actually matters: is this lab actually okay?
Report Data Sources
- NetBox, devices, IPs, VLANs, services
- Proxmox, VM/container inventory
- Docker, running stack list
- Restic, backup job status
- SMART, disk health
- Uptime Kuma, service availability
- apt/OS, update status
Report Sections
$2. Backup health
$3. Disk health
$4. Service availability
$5. Security findings
$6. Changes since last report
$7. Action items
Start Simple
A Bash or Python script, triggered by a cron job or systemd timer, that pulls status from each tool's CLI/API and writes a Markdown report, then emails or posts it to Discord.
Level Up With n8n
- n8n orchestrates the collection on a schedule
- NetBox API supplies the inventory
- The
resticcommand supplies backup results - The Uptime Kuma API supplies health status
- The finished report gets saved into Gitea or Paperless-ngx
Redaction
- Never include secrets in the report
- Avoid publishing public IPs or internal network details in any notification channel that leaves your network
Human Review Stays in the Loop
Automation generates the report. A human decides what to actually change, this capstone is about visibility, not autonomous remediation.
Security & Backup Notes
- This report is a map of your entire infrastructure, store and transmit it with the same care you'd give the NetBox instance itself
Troubleshooting
- APIs lack permission, check the API tokens used have read access to the specific data being pulled
- Report exposes sensitive data, add redaction rules before this report ever leaves your network
- Automation silently fails, add a "did the report actually run" check, not just trust in the schedule
- Data sources disagree, reconcile discrepancies (e.g. NetBox says a host exists, Uptime Kuma has no monitor for it) as an action item, not an error to ignore
- Alert fatigue from unchanged warnings, surface only new or worsening findings prominently, not the same stable warning every week
Lab Finish Line
A weekly report that identifies at least one service, one backup status, one disk-health status, and one action item.