TrueNAS Scale homelab: recovery without SSH or console

Yesterday I had one of those "oh no" homelab moments that makes your stomach drop. I'd just finished transferring several hundred gigabytes — over a million files — to my TrueNAS Scale box. Walked away for the day, came back… and both my Mac mini and MacBook couldn't reach the web UI. SMB shares in Finder dropped. Reboots didn't help. I plugged a monitor straight into the NAS, rebooted via console a few times, then made the classic mistake: hot-swapping the Thunderbolt splitter too aggressively. That completely corrupted the Thunderbolt handshake. Suddenly I had zero console access (HDMI or TB), the dashboard was still dead, and I was staring at a brick with my data inside. Worst part? I had never enabled SSH. So my only two doors in were the broken web UI and the now-dead physical console. Temporary lockout achieved.

That's when ClawQL + editor agent became the real MVP. Instead of jumping between random chat windows and losing my train of thought, I used ClawQL's memory_ingest to build a living, timestamped case study right inside editor agent. Every hypothesis, every command, every log, every decision got automatically saved and recalled. I could ask it "what did we try with the routing table?" and instantly get the full context back. Gemini gave a couple of ideas, but ClawQL kept everything organized and evidence-based so I never chased ghosts.

Within an hour I had the Mac's routing table cleaned up (hello, nine utun interfaces from Docker Desktop + dual 192.168.0.x IPs on the same subnet), web UI restored, SSH enabled, and the NAS fully debugged from the terminal. No data loss. Pools stayed ONLINE. Everything back online.

I turned the entire incident into a detailed case study that walks through the outage, every troubleshooting step, the exact ClawQL workflow I used, and the lessons learned so I never repeat them.

Biggest takeaways I'm implementing today

  • Enable SSH + auto-start on every NAS the day you set it up (not the day it breaks)
  • Never trust hot-swapping Thunderbolt for console duty
  • Keep client networking clean — Docker/K8s utun interfaces can silently murder connectivity
  • Real-time documentation isn't optional; it's your safety net

ClawQL turned a potential multi-hour (or multi-day) panic into a structured, calm recovery and left me with a permanent, searchable record of exactly what worked. If you run homelabs, self-hosted infra, or anything where downtime hurts, you need this in your toolkit.

Would love to hear your own "I thought I was locked out forever" stories in the comments. What saved your last homelab crisis?

Technical deep-dive

The long-form write-up with commands, timelines, and appendices lives in the ClawQL docs: use the Case study (docs) link below.

Original thread (comments, reactions, and any media): View on LinkedIn

Related links