Linux commands for beginners

Simple Commands

agent-screen — show Hive client agent (you can exit from it with Ctrl+A, D)

firstrun -f — ask for rig ID and password again

mc — file manager like Norton Commander but for Linux

selfupgrade — upgrade from console, same as hitting a button on the web

sreboot — do a hard reboot

sreboot shutdown — hard shutdown

Miners

miner — show running miner screen (you can exit from it with Ctrl+A, D)

miner start, miner stop — start or stop currently configured miner obviously

miner log, miner config — selfexplaining

System Logs

dmesg — to see system messages, mainly to see boot log

tail -n 100 /var/log/syslog — to show last 100 lines from system log

Network

ifconfig — show network interfaces

iwconfig — show wireless adapters

Keyboard Shortcuts

Ctrl+C — stop any running command

Switch miners screens, detach from terminal:

Ctrl+A, D — detach from screen (miner or agent) to leave it working

Ctrl+A, Space or Ctrl+A, 1,2,3 — switch between screens if you have second miner running and so on

Advanced Commands

Status / Diagnose

agent-screen log — show logs of various parts (you can try log1 and log2) of the Hive agent

hello — say hello to server: to refresh IP addresses, configs etc. (normally it’s run at startup)

net-test — check and diagnose your network connection

timedatectl — show time and date synchronization settings

top -b -n 1 — show list of all processes

wd status — show hashrate watchdog status and log

AMD

amd-info — show AMD cards info

amdcovc — show info about AMD cards power supply

amdmeminfo — show AMD cards memory info

wolfamdctrl -i 0 --show-voltage — show voltage table for AMD GPU #0

Nvidia

journalctl -p err | grep NVRM — show recent Nvidia GPU errors, if any

nvidia-info — show extended Nvidia cards info

nvidia-driver-update — update Nvidia drivers

nvidia-driver-update 430 — download and install latest driver from series 430.*

nvidia-driver-update --nvs — reinstall nvidia-settings only

nvidia-smi — show Nvidia cards info

nvtool --clocks — show core/mem clocks for all the Nvidia GPUs

Hardware

gpu-fans-find — spin GPU fans from the first card to the last to make it easier to find the required GPU. You can specify the card number, then the fans will start on it. If the worker starts up correctly, it's quite easy to track which card corresponds to which bus. For example, if the GPU0 card has a 01:00.0 bus and you want to find it, use the command gpu-fans-find 0. This command will spin up the fans only on this card for a while. And so on, following the example.

sensors — show voltage/temperature readings of the motherboard and CPU

sreboot wakealarm 120 — shutdown PSU and boot in 120 seconds

/hive/opt/opendev/watchdog-opendev power — send a power command to OpenDev watchdog

/hive/opt/opendev/watchdog-opendev reset — send a reset command to OpenDev watchdog

Upgrade / Install

disk-expand -s — expand a Linux partition to fill remaining drive space

hpkg list miners — list all installed miners

hpkg remove miners — uninstall all miners

nvidia-driver-update --remove — remove all downloaded Nvidia driver packages except currently installed

selfupgrade --force — force upgrade; it can help in situations when selfupgrade says Hive is up to date but actually it isn’t

Logs

journalctl -u hive --no-pager — show Hive service boot log

journalctl -u hivex --no-pager — show log of X server (graphical user interface)

logs-on — write all logs to disk, they will remain after reboots

logs-off — write all logs to RAM to reduсe USB flash drive wear (default)

log='/var/log/syslog'; gzip -c9 "$log" | base64 -w 0 | message file "$(basename "$log")" payload — send /var/log/syslog file to dashboard