Running within a Docker, t-rex miner dials to load with "WARN: NVML error, code 12"
Output:
20211025 16:05:08 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...
20211025 16:05:09 T-Rex NVIDIA GPU miner v0.24.5 - [Linux]
20211025 16:05:09 r.3ed63f02e8cb
20211025 16:05:09
20211025 16:05:09 WARN: Can't load NVML library, dlopen(25): failed to load libnvidia-ml.so, libnvidia-ml.so: cannot open shared object file: No such file or directory
20211025 16:05:09 WARN: NVML error, code 12
20211025 16:05:09 WARN: Can't initialize NVML. GPU monitoring will be disabled.
20211025 16:05:09
20211025 16:05:09 NVIDIA Driver version N/A
20211025 16:05:09
20211025 16:05:09 + GPU #0: [00:2d.0|2489] GeForce RTX 3060 Ti, 7981 MB
20211025 16:05:09
Workaround I implemented:
After starting docker containing with bash shell>connected to shell>'ln -s /usr/lib/x86_64-linux-gnu/libnvidia-ml.so.465.31 /usr/lib/x86_64-linux-gnu/libnvidia-ml.so'
There was a symbolic link 'libnvidia-ml.so.1 -> libnvidia-ml.so.465.3' but no 'libnvidia-ml.so' link, this was the result after the workaround:
root@959faa8fc6eb:/trex# ls -la /usr/lib/x86_64-linux-gnu/ | grep -i libnvidia-ml
lrwxrwxrwx 1 root root 22 Oct 25 16:54 libnvidia-ml.so -> libnvidia-ml.so.465.31
lrwxrwxrwx 1 root root 22 Oct 25 16:54 libnvidia-ml.so.1 -> libnvidia-ml.so.465.31
-rwxr-xr-x 1 root root 1774808 May 18 14:18 libnvidia-ml.so.465.31
I don't know if this is an issue with the docker image 'ptrfrll/nv-docker-trex' I pulled into Unraid or with the T-Rex miner docker image, so I am just posting inform others and trying to see if this is a bug.