Metrics → weather DEMO
Scenarios
Node
Every piece is a resource: CPU → chips · RAM → DIMMs · disk I/O → NVMe · network → RJ45 · GPU → cards.
The pile is history: sustained load piles up; it melts away when the node rests.
Heat (hottest of CPU and GPU): >60 °C the pile glows red and melts slower · ≥85 °C it stops melting, shakes and smokes.
Trouble: pool not ONLINE or disk errors → a broken HDD that never melts + a storm · scrub/resilver → fog · RAM >90% → DIMMs slide off · packet loss → tangled cables.
Money: coins rain at each node's power draw (from your UPS where you have one, otherwise GPU + CPU + a base figure) × your price per kWh.
Lings: one little gremlin per 30 days of node uptime (max 4 per node). Reboot and they run away. The demo runs uptime fast.