Understand the fundamentals
Read identity and state before the big number
Before you look at the big hashrate number, check which machine you are looking at. Note the asset, the reported model and the firmware, which is the miner’s installed software. Also note the active profile, the set of operating settings it is using. Then check what the machine is doing. It may be starting up, tuning its settings, hashing, stopped on purpose or showing an alarm. Uptime usually shows the time since the last restart, or since the miner’s software last started. Check what uptime means on this particular screen.
A dashboard gathers readings from different sources onto one screen. Some values update often. Some are averages over a period of time. Some can stay on the screen after the miner stops communicating. Before you trust the screen as current, check the timestamp or when it last refreshed.
See how it works
Connect the fields to the parts
Hashrate is how fast the miner is computing. Chain or board fields show the groups of chips that the firmware reports. A temperature field shows the reading from one labeled sensor, or an estimate. Fan readings show how fast each measured fan spins. Power values may be measured or estimated, depending on the equipment. Error counters add up events over a set period, or since the counter was last reset.
Each field answers only one small question. Seeing the expected number of chips does not prove the pool is accepting every share. A normal temperature does not prove the miner is set up for the right pool. To describe the whole machine, you need several readings that agree with each other.
Read the labels around the big number
Swipe sideways to follow the full diagram →
| Field | Meaning | Common interpretation error |
|---|---|---|
| Hashrate | Computing rate or estimate | Treating it as guaranteed revenue |
| Fan RPM | Rotation speed (revolutions per minute) | Assuming it measures airflow |
| Temperature | Reading from one sensor, or an estimate | Comparing different sensor locations |
| Uptime | Time since the last restart or software start | Assuming no earlier interruptions |
| Error counter | Running total of events | Ignoring the time period or counter resets |
Connect it to the field
Worked example: a counter without a time window
A dashboard shows 100 errors. Is that a lot? One hundred errors in one minute is very different from one hundred over six months. A reboot may also reset the counter. Before you compare it with another machine, record the starting count, how much time has passed and anything else that changed.
A useful note reads “counter increased from 100 to 110 between 09:00 and 09:10.” That is ten new events in ten minutes. Compare that with “ten errors means bad board.” The first note keeps the evidence. The second jumps to a conclusion. It skips what the counter means, the time period and the checks needed to support a cause.
Terms used in this lesson
- Value and unit
- What is measured and how it is expressed.
- Time window
- Whether a value is a reading at one moment, an average over a period or a running total.
- Status context
- The mode, uptime and alarms that help explain the reading.
Optional references & further reading
Use current manufacturer instructions for the exact equipment and your site’s approved procedures. Course levels do not grant task authorization.