ccminer Universal NVIDIA Docker Miner

[Patrick]

Here is the docker image for ccminer. The docker image is built with CUDA 9.0 from the official nvidia/cuda image so you need to have CUDA 9 compatible drivers installed on the host system. You do NOT need to have CUDA installed on the host system.

This is also meant for nvidia-docker2 so we use the docker runtime flag:

docker run --runtime=nvidia -it -e algo=<pick your algorithm> -e pool=<IP or DNS address of the pool> -e startport=<pool port> -e intensity=<default 27 but can change this> -e username=<wallet or ID> -e pass=<x or pass for pool> servethehome/universal_ccminer:latest

If you want to run skien, blake2s, nist5, monero, zcash you name it, just pick the algorithm there.

This image clones the ccminer application, sets up the build, and then runs the application based on the variables you use at the runtime.

 

[hifijames]

Does it require the install of nvidia-docker2 first?

 

[hifijames]

I had to remove '--runtime=nvidia' so it will pull the docker image.
But then I got this:
./ccminer: error while loading shared libraries: libcuda.so.1: cannot open shared object file: No such file or directory

Adding back the runtime variable gives error 'Unknown runtime specified nvidia'.
This CUDA thing always confuses me.

 

[Patrick]

@hifijames you should need nvidia-docker2. Built the new image looking at the currently supported nvidia-docker. I do know it is a pain since I had been running the old nvidia-docker and had to re-do.