Miner status & chip state
Miner status badge
The Status column in Machines and the badge on the miner page are a human reading of the last scan plus any command the Mole agent is still carrying out. The rule, in order:
| Badge | Colour | Meaning | Computed from |
|---|---|---|---|
| Resuming… | cyan, pulsing | A start / wake / power-mode command is queued or being executed | an in-flight resume_mining, wake, start_mining, power_on or power_mode command |
| Stopping… | orange, pulsing | A stop / sleep command is queued or running | in-flight stop_mining, sleep, pause_mining, power_off |
| Rebooting… | violet, pulsing | A reboot is queued or running | in-flight reboot, restart, reset |
| Updating… | violet, pulsing | A firmware install is queued or running | in-flight install_firmware / firmware_update |
| Offline | red | The agent could not reach the miner | scan status offline, dead, inactive or off |
| Error | amber | The miner answers but reports an error (for example Can not connect to pool) | scan status error/failing, or reachable with a pool error |
| Stopped | grey | Sleep / low-power mode | power mode sleep or scan status stopped |
| Mining | green | Reachable and hashing | status active/online/healthy with hashrate > 0 |
| Starting | sky blue | Reachable, hashrate ramping up | status active with a "lower than ideal / warming / ramp" message |
| Online | sky blue | Reachable but not hashing (no error reported) | status active with 0 hashrate |
| Unknown | grey | Never scanned or unrecognised status | anything else |
The badge tooltip shows the detail (the error text, or "queued for 14:32:05" for a staggered command). In-flight commands come from GET /miners/in-flight and are refreshed with the Machines table.
Chip state
Since agent v3.285.0 each hashboard reports its working chips and the model's expected count. The Chips column (Machines) and the Hashboards & chips card (miner page) grade them:
| State | Label | Rule |
|---|---|---|
| ok | All chips | working chips = expected chips |
| degraded | Chips lost | 0 < working < expected |
| dead | No chips | board detected, 0 chips |
| off | Board off | 0 chips and the firmware reports the board as inactive |
| missing | Board missing | the control board does not see the slot at all |
| unknown | Not reported | the agent has not sent board data for this miner |
Per miner, the summary is dead if every board is dead/missing, degraded if any board is degraded, dead or missing, otherwise ok. The ratio bar is working / expected chips; when the agent sends no per-board expectation, the miner's expected total is split evenly across its boards.
Older agents
An agent older than v3.285.0 only sends summary figures. Those miners show Not reported (–) in the Chips column and "The agent has not reported this miner's boards yet" on the miner page. Nothing is estimated in their place.
Lifecycle state
Separate from the live status, each miner has a State set by people (Edit miner → Miner options, or the Repair console):
| State | Meaning |
|---|---|
| Deployed | In a rack, expected to mine — issues and alerts apply |
| In repair | Taken out for repair; most issue rules are muted |
| Spare | On the shelf, not expected online |
| Decommissioned | Retired; kept for history |