Closed Singman33 closed 10 years ago
Any time a GPU goes sick once the miner shuts down the cl devices are locked and the whole thing becomes a mess and you cannot restart the miner without restarting the machine.. 99 of 100 times when this scenario happens it is due to bad user configuration. Try using the develop branch and see if it cures your issue.
I'm often switching algorithms, not by using the pool feature, but by switching down the miner and relaunching with a different config file. If I'm doing that every 15 minutes, the miner will hang when shuting down and block GPUs, so the next instance can't work. That happend very very often with Scrypt, X11, X13 algorithms That dont happen when using Scrypt with a different miner (let say, cgminer or previous version of sgminer). I'm afraid that you (devs) have broken totally this miner by trying to improve it.
'# ps aux | grep miner root 4556 76.3 0.0 0 0 ? Zsl 16:23 9:17 [sgminer-v5]
root 5656 0.0 0.0 27276 1620 ? Ss 16:33 0:00 SCREEN -dmS MINER sgminer-v5 --config /home/eric/globalboost.conf
root 5657 0.0 0.2 139988 13948 pts/2 Ds+ 16:33 0:00 sgminer-v5 --config /home/eric/globalboost.conf