I've been testing the version 3 yesterday and I have an issue to report. It might be something I overlooked because of the new design, but let me start by explaining how I work the version 2 atm.
I have 2 different kinds of cards on my rig, so each kind of card is more profitable in different algos. Because of that, I don't want to run an average of all my cards on every algo and instead I've made 2 folders of NemosMiner, each folder with different data inside the "stats" folder and I run 2 instances at the same time, one for each group of cards.
At first the benchmark would get confused when two ccminers were running and it'd update the hashrate incorrectly. In order to fix that, I run the benchmark on the 1st group and then turned the "*HashRate.txt" files into read-only. I then benchmarked the 2nd group of cards and again locked their hashrate files into read-only. Since then, running 2 instances at the same time works like charm, the only drawback is that the hashrates are not being updated automatically but if you don't change anything in your system, the hashrate rarely changes because of an updated miner... and of course you can just run the benchmark again!
I believe the above trick would help people with more cards than the miner can handle? Maybe?
Anyway to be honest I posted the above long explanation in case someone wants to try it, the issue I have with version 3 with the above setup is pretty simple... I can't setup specific cards for each instance OR I'm doing it wrong and there's no documentary to help me!
You could setup if you switch to manual edition of the config files.
However, NPlusMiner now supports running multiple instances. I guess that might ease your life
https://github.com/MrPlusGH/NPlusMiner/releasesMultiple instances doesn't fully work for me. If i start the second after the first instance is running, the second instance would close the miner for the first instance until the sleep cycle is over.
Also if i do not make the hashrate files read-only, there is a chance that it will get overwritten.
What i have done for now is to benchmark separately, making the hash-rate files read-only, then starting both instances simultaneously.
The other thing i noticed is, and i'm not sure if its a bug, is i have about 10 different instances running at the same time. All 10 would have different btc/gh/day values. The difference is about a couple of percent, but this means on the same rig with the same cards, i would have two instances mining different algos. Is this normal?
