0

I have a mining rig with EthOS on it. I would like to send all informations like hash rate, temperature, etc to my server. My question is how can I catch all the information from the rig in order to send it to my server?

Thanks in advace

1
  • Thats sounds greate. I will apply it in my server. Thanks a lot. – toye Dec 27 '17 at 23:37
0

At my mining farms I'm using a Zabbix server with agents installed on our rigs which are running Ubuntu 16.04.3LTS. You can do the same using EthOS since it is a Debian based distribution if I recall. This will allow you to easily send data from your rigs back to your Zabbix server which can then graph the data into very nice looking charts.

You can also take this a step further and setup alerts so that you can get an email notification if one of your rigs goes offline, or your stale share rate goes above a certain threshold, text message alerts, telegram alerts and so on.

I'm using bash scripts I wrote to retrieve data from syslog, but you could also write python or perl scripts or whatever language you're comfortable with. Zabbix also allows you to do remote command execution so you could automate some disaster response situations such as remotely restarting software daemons that might've crashed.

https://www.zabbix.com/

3
  • I've tried to run zabbix-agent on my EthOS Client. After editting the zabbix-agentd.conf file it is not possible to run the agent. After reinstalling the agent is running automatically. Do you know what I am making wrong? I installed the agent on my zabbix server. There it runs without any problem and I can see the reports on the zabbix frontend. – toye Jan 8 '18 at 15:03
  • I imagine it's a syntax error of some sort. You probably mistyped an option. Try running the zabbix agent after you make the change, then checking /var/log/syslog or /var/log/zabbix_agentd.log – hextet Jan 9 '18 at 7:22
  • When the agent can not be startet then the log file is empty. It just do nothing. I also tried only to change the server ip from localhost to my server's ip. The agent also doesn't run. Without a log file it is really hard to find the issue. – toye Jan 9 '18 at 8:21
1

After we build a small farm with 50 rigs (~500 GPUS) a lot of problems came, like: gpu temperature, network failure, self reboot, room heat and so on. As a solution we start building our own monitoring software with a simple idea:

  • every rig will run a small client
  • a VPS (outside farm) keep all rigs connected and communicate with them realtime throw a TCP socket connection
  • on VPS a simple User Interface to manage, see, change things on selected rig

We also made a simple website for the software, can be found here: https://rig-manager.com/

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.