Hashrate Watchdog Issues

#1

So, I had to do some work on some of my miners the other day. All running 0.6-28@190408. I issued the ‘miner stop’ command and I started my work. To my surprise the Hashrate Watchdog was still loaded and checking to see if the miner was running. The miner started again and I killed it. Then a few minutes later my rig rebooted and I lost everything that I was doing.

Is there a way to stop or unload the Hashrate Watchdog when you are doing maintenance to the system. I read in an earlier post that this had been fixed back in June 2018, but I guess it hasn’t been tested in a while.

#2

In the top right corner. Select the hashrate watchdog and toggle it off. Then be sure to scroll down and hit the save button. This will stop the hashrate watchdog from sending a reboot command.

#3

Thanks for the reply. I am aware of how to enable and disable the hashrate watchdog. However, that is not why I have created this thread.

The watchdog is there to handle unexpected issues that may happen during the course of mining. Issuing the ‘miner stop’ command is not one of these issues and is expected behavior. The watchdog should have the ability to be paused if this command is issued and from what I have read, that functionality used to be there. If the watchdog is not currently programed to do so, I would recommend that as a feature improvement.

My bigger question is from the Rigs perspective, how do I interact with the watchdog? Example, I am on CLI and I want to disable the watchdog for a period of time while I am doing maintenance. How do I go about doing this? I have been through the /hive/bin/miner and /hive/bin/miner-run scripts and there is nothing there that calls the watchdog. This leads me to believe that the watchdog is actually being done on the hive server side and a miner restart and reboot command are being sent from the server. If this is the case, can I pause the watchdog or toggle it on and off via CLI?

#4

To start and stop it from CLI I would use:

systemctl stop hive-watchdog.service will stop it
systemctl start hive-watchdog.service will start it

#5

I think I have found where the watchdog is. The watchdog script is located in /hive/bin/wd. It has a few arguments that can be given here is a list.

wd run|start|stop|restart|log|status

The start|stop|restart arguments seem to be the most important ones that interact with systemd to control the service.

The ‘wd’ script is being called from the ‘hive’ script towards the end of the script. Here is the IF statement that is kicking it off.

if [[ $MAINTENANCE -eq 1 ]]; then
        echo2 "> Maintenance mode. Miner and software watchdog are disabled"
else
        echo2 "> Starting miner"
        miner start



        #Start Watchdog. It will exit if its 0
        [[ $WD_ENABLED == 1 ]] &&
                echo2 "> Hashrate Watchdog enabled" &&
                wd start
fi

Interestingly, it looks like there is a way to enter maintenance mode on the Rig. I will have to play around with this and report back. Since the devs are mostly unresponsive, I guess I just have to resort to reverse engineering their code.

#6

Thanks for the reply. That will work for sure, but now I am interested in how to enter into maintenance mode, per the code spinet that I posted.

1 Like
#7

Nice find!
You need to find the part of the code that sets the value for $maintenance
The script seems to treat it as a Boolean. 1 = Maintenance mode on and 2 = Maintenance mode off

#8

Yep, I found it, but it looks like you have to reboot the rig to enter into maintenance mode, so I will just stick with issuing the ‘wd stop’ command for right now.

For future reference, it is in the $RIG_CONF or /hive-config/rig.conf file.

1 Like