Having issues with my U3 still. Having to reset R1 everytime I come home from work to get the U3 to start back up. The R1 is hashing away fine and wifi working great but the U3 keeps dying.
Im using the standard U3 settings. Any recommended settings to keep it running that has worked for others? (volts, etc..)
The R1 is great at recovering from zombies with a cgminer restart(via the init.d script) and in some instances it requires a reboot. I have never had to replug my U3 in when using the R1 as a controller. The problem is, the R1 has no script monitoring for U3 zombies and restarting cgminer. I've taken one step to resolve this issue and so far it's kept my U3 running for 2 days.
Add the following to the end of your cgminer-monitor script(/usr/bin/cgminer-monitor), and let me know how it works out for you:
numU3MinersInactive=`/usr/bin/cgminer-api devs localhost 4029 | grep "\[No Device\] => true" | wc -l`
if [ "$numU3MinersInactive" -gt "0" ]; then
echo "$numU3MinersInactive miners have gone inactive, restarting cgminer-u3"
/etc/init.d/cgminer-u3 stop
sleep 5
/etc/init.d/cgminer-u3 start
sleep 20
noASCMsgCount=`/usr/bin/cgminer-api devs localhost 4029 | grep "\[Msg\] => No ASCs" | wc -l`
if [ "$noASCMsgCount" -gt "0" ]; then
echo "No Miners Detected, trying one more cgminer restart"
/etc/init.d/cgminer-u3 stop
sleep 5
/etc/init.d/cgminer-u3 start
sleep 20
noASCMsgCount=`/usr/bin/cgminer-api devs localhost 4029 | grep "\[Msg\] => No ASCs" | wc -l`
if [ "$noASCMsgCount" -gt "0" ]; then
currentDate=`date`
echo "$currentDate - No Miners Detected after 2 restarts, rebooting" > /tmp/cgminer-u3_reboot.log
reboot
fi
fi
fi
Brief description of what cgminer-monitor and this snippet are doing:
- cgminer-monitor runs every 3 minutes via cron. Its main purpose was to keep the integrated R1 ASIC up and running.
- The snippet uses cgminer-api to get a list of inactive miners under the cgminer-u3 instance.
- If an inactive(zombie) miner is detected, cgminer is restarted.
- If cgminer comes back up with no miners attached, it tries one more restart
- After second restart, if no miners are attached, it reboots the R1.
