xmrig docker container

This container allows running xmrig in a docker container.

Usage

• A simple one liner to get the container mining right away

docker run -it -d --name xmrig patsissons/xmrig -o ca.minexmr.com:5555 -u 47NHecs6qjvDcbx3eW6cDGDwdm3gDqbHs7G8hzPYRxf3YRTcDJw8kXhDxfHinsjHUwVwdFusSn76UHkaz68KurUgHvFmPMH.github-xmrig -p x -a cn/r

• A more comprehensive script to simplify the process of starting and uprading the miner. Running sudo ./start-xmrig.sh will pull down the latest version while still mining, then stop and upgrade the container to resume mining with minimal downtime. If you want to use this script with the latest develop builds, run sudo ./start-xmrig.sh develop instead. The script is ready to accept overrides for a few environment variables to simplify multi-host distribution.

Environment Variable	Default Value	Notes
ADDRESS	N/A	make sure to replace the default with your own address or you will be mining to my address
HOST	hostname -s	override this if you want a custom host name that is different from your actual host name
PORT	8080	if another container or process is already using port 8080 this can be adjusted
POOL_HOST	N/A	make sure to replace this with your own pool remote uri
POOL_USER	$ADDRESS.$HOST-$TAG	adjust this to your needs
POOL_PASS	x	adjust this to your needs
POOL_ALGO	cn/r	algorithm to use, see docs

Host configuration

Everything has been tested on Ubuntu 16 LTS (Xenial)

AMD requirements

AMD requires only drivers installed on the host to interact with the hardware.

• install the AMDGPU-Pro drivers (17.40-514569) on the host machine and then run the docker image with --device=/dev/dri --device=/dev/kfd --group-add=video.

You can verify that everything is working on the host with the following commands

• check the status of amdgpu-pro by running /opt/amdgpu-pro/bin/clinfo
• test the container integration by running docker run -it --rm --device=/dev/dri --device=/dev/kfd --group-add=video --entrypoint=/bin/bash patsissons/xmrig -c /opt/amdgpu-pro/bin/clinfo

NVIDIA requirements

NVIDIA requires both a docker runtime and cuda drivers installed on the host to interact with the hardware.

• Install the Ubuntu 16.04 cuda-drivers (10.2) (use apt-get install --no-install-recommends nvidia-container-toolkit to install)
• Install the nvidia-docker runtime (use apt-get install --no-install-recommends cuda-drivers to install)

You can verify that everythign is working on the host with the following commands

• check the status of the cuda-drivers by running /usr/bin/nvidia-smi
• test the container integration by running docker run -it --rm --runtime=nvidia --entrypoint=/bin/bash patsissons/xmrig -c /usr/bin/nvidia-smi

Building

This image uses a multi-stage build, all that should be required is docker build . to build the latest image.

Environment

You can adjust the following docker environment variables to customize the build process.

• GIT_REPOSITORY: use a custom git repo for xmrig (default is https://github.com/xmrig/xmrig.git)
• GIT_BRANCH: use a custom branch or tag
• CMAKE_FLAGS: use custom cmake flags
• PACKAGE_DEPS: use custom installation packages

You can adjust the following docker build args

• CUDA_VERSION: CUDA container version (default is 10.2)
• CUDA_UBUNTU_VERSION: Ubuntu version of CUDA container (default is 16.04)
• AMDGPU_VERSION: AMD driver version (default is 17.40-514569)
• ALPINE_VERSION: image to use for alpine minimalist container (default is latest)

Donations

If you find this docker image useful, donations for work on dockerizing the build and mining app are appreciated at any of the folllowing addresses:

• BTC: 1LNY9wSPs913Y9jXMTrrVze1E41nhm6Qv7
• LTC: LhnwdbrnQaQbjDkqxXFmxXGPcFhMBA9gFu
• XMR: 47NHecs6qjvDcbx3eW6cDGDwdm3gDqbHs7G8hzPYRxf3YRTcDJw8kXhDxfHinsjHUwVwdFusSn76UHkaz68KurUgHvFmPMH