Just saw this post. To add on its a lot easier to just run Flexfarmer so there’s no need for the full node or syncing.
Hive OS runs on Linux so its possible to farm Chia using Flexfarmer while GPU mining. The link to the guide is below, check the Flexfarmer reddit post for additional details.
Note that this guide does require some familiarity with Linux and we do not officially support this as its a project done up by a community member. Hopefully someone will do up a better guide/video.
@Br4ne I have a similar set up. The official guide on farming on multiple machines is thorough.
The rest is just familiarizing yourself with using a cli. This guide also helped me regarding mounting drives automatically.
The only issue I’ve had is getting crontab working.
Hi, I have my hiveos on a 1 tb hdd. Can you explain every step in detail? Please.
One last thing: How to update your local chia installation.
First go to the chia directory and activate venv:
cd /home/chia-blockchain
. ./activate
next stop all chia services:
chia stop -d all
after that fetch all changes and start the update:
deactivate
git fetch
git checkout latest
git reset --hard FETCH_HEAD
chmod 777 install.sh
./install.sh
. ./activate
chia init
After that start the chia services again:
chia start farmer
After latest change I hat problems with permissions of the config folder, you can fix that with:
chia init --fix-ssl-permissions
Plotting on your hive rig really a bad idea. The plotting impacts the hive os. I tried doiing this with on a rig with intel i7 9700 with 32gb and 1tb nvme drive and sending finished plots to a hard drives on another computer using SMB and 1g Ethernet. I ended up with high LA issues. This is expected becuase all the CPU time is plotting but hive will see it and if you have watchdog enabled, your machine will reboot over and over again as the plotting ramps up. This totally disruped my GPU mining. Also transferring plots across network will cause high LA if you have slow storage on the remote end.
After that, I rethought my process and attached a 12bay storage chassis to my hive rig and did the plotting on on other computers and then transferred the full drives to the storage chassis. I set up chia as a harvester reporting to chia Full node on my Main PC. Don’t need as much CPU resources to run a just a harvester.
Let’s say… I just wanna Harvester on my HiveOS.
Could you advise me how to get CA import, and else to get it running?
(I would run full node on my PC, and it’s more visual just in case)
Already all explained here:
I have my chia setup on windows at the moment, but i want to change system to hive os. How do i activate my hdds full of plots in chia? Can i use all other forks as well, for example flax, hdd, staicoin?
This crontab entry seems to do nothing.
I do not know why the start node at start-up did not work. This is what I have done. I wrote a short bash script that checks if the node is running and start it if the node is not. Then added the cronjob to run every 20 minutes.
This is the body of check_running.sh
chia_pid=`pgrep -f chia_farmer`
if [[ $chia_pid -gt 1 ]]
then
echo "`date` Check Successful PID: $chia_pid"
else
echo "`date` Starting Farmer"
cd /usr/lib/chia-blockchain ; . ./activate ; chia start farmer
fi
do not forget to set the script as executable:
chmod u=xrw check_running.sh
Then in /hive/etc/crontab.root add the following:
# Check Chia node status
*/20 * * * * {/path/to/script}/check_running.sh >> /var/log/chia_checks.log
Assuming you have the path correct on your own script this will log the checks to /var/log/chia_checks.log. Once you have fully verified the whole system is working as intended we might consider removing the logging.
Update: changed the bash script because the if statement was reading the empty PID as non-zero
I also found that the corn job was running in the root environment so I added a symlink to the .chia directory. ln -s {/path/to/data}.chia /root/.chia
So cronjobs simply do not work for chia some reason even if you get them to apply correctly. Might have something to do with venv. Creating a service works but HiveOS seems to overwrite services in systemd regularly. Anyone familiar with how to make a service in HiveOS that won’t overwrite on update or reboot?
This topic was automatically closed 416 days after the last reply. New replies are no longer allowed.