Welcome to Cerberus Discussions!

[CTOUT]

👋 Welcome to Cerberus!

Cerberus is a CPU thermal watchdog for TrueNAS SCALE. It runs every minute via cron and progressively stops Docker/TrueNAS apps during thermal events — busiest CPU consumer first — then restarts them in dependency order once the system cools.

This Discussions space is for everything that doesn't belong in an issue: questions, ideas, hardware reports, and community conversation.

---

What this space is for

• ❓ Questions — stuck on installation, threshold tuning, the cron setup, or recovery ordering? Ask here.
• 💡 Ideas — suggestions for new features, additional thermal driver support, or configuration improvements
• 🖥️ Hardware reports — running Cerberus on hardware other than the HPE Gen10? Share your sensors -j output and help expand platform support
• ��️ Roadmap feedback — the TODO.md tracks what's planned; input on priorities is welcome
• 🤝 Getting started help — if something isn't clear in the README, ask here and the answer will improve the docs

---

What's on the roadmap?

The current priority is:

1. Broader thermal sensor support — get_temp() is currently hardcoded to the k10temp driver (AMD Opteron X3000-series). Adding support for coretemp (Intel), zenpower (AMD Zen alternative), and nct6775/nct6779 (Nuvoton Super I/O) requires live sensors -j output from those platforms.

If you're running Cerberus on hardware other than the HPE ProLiant MicroServer Gen10, please open an issue or drop a message here with the output of:

sensors -j 2>/dev/null | jq 'keys'
sensors -j 2>/dev/null | jq 'to_entries[] | {key: .key, fields: (.value | keys)}'

Full details in TODO.md.

---

Quick links

• 📖 README — installation, configuration, and how Cerberus works
• 🗺️ Roadmap (TODO.md) — tracked expansion items
• 🔒 Security policy — how to report vulnerabilities
• 📝 Contributing — how to contribute

---

To kick things off — drop a comment with your hardware and what TrueNAS apps you are protecting with Cerberus. 👇