lad1337 / netdata-glibc

netdata with glibc package for use with nvidia-docker2

netdata-glibc

This is an automated build of netdata with glibc package for use with nvidia-docker2 / nvidia-container-toolkit. Now available in Unraid Community Applications.

Netdata with Nvidia GPU monitoring in a container. This image was created due to netdata/netdata using Alpine, a musl distribution, as a base. Nvidia Docker / Nvidia Container Toolkit is only compatible with glibc distributions. This image uses netdata/netdata as a base and adds a GNU C library to run binaries linked against glibc. This image does not contain nvidia-smi, but is compatible with nvidia-docker2, nvidia-container-toolkit and the Unraid Nvidia Plugin.

Docker 19.03 + nvidia-container-toolkit

docker run -d --name=netdata \
  -p 19999:19999 \
  -v /proc:/host/proc:ro \
  -v /sys:/host/sys:ro \
  -v /var/run/docker.sock:/var/run/docker.sock:ro \
  -e PGID=<HOST_DOCKER_PGID> \
  -e DO_NOT_TRACK= \
  --gpus all \
  --cap-add SYS_PTRACE \
  --security-opt apparmor=unconfined \
  d34dc3n73r/netdata-glibc

< Docker 19.03 + nvidia-docker2

docker run -d --name=netdata \
  -p 19999:19999 \
  -v /proc:/host/proc:ro \
  -v /sys:/host/sys:ro \
  -v /var/run/docker.sock:/var/run/docker.sock:ro \
  -e PGID=<HOST_DOCKER_PGID> \
  -e NVIDIA_VISIBLE_DEVICES=all \
  -e DO_NOT_TRACK= \
  --cap-add SYS_PTRACE \
  --security-opt apparmor=unconfined \
  d34dc3n73r/netdata-glibc

docker-compose

version: '3'
services:
    netdata:
        container_name: netdata
        image: d34dc3n73r/netdata-glibc
        ports:
            - 19999:19999
        restart: unless-stopped
        environment:
            - NVIDIA_VISIBLE_DEVICES=all
            - DOCKER_HOST=proxy:2375
            - DO_NOT_TRACK=
        cap_add:
            - SYS_PTRACE
        security_opt:
            - apparmor:unconfined
        volumes:
            - /proc:/host/proc:ro
            - /sys:/host/sys:ro
    proxy:
       container_name: proxy
       image: tecnativa/docker-socket-proxy
       volumes:
            - /var/run/docker.sock:/var/run/docker.sock:ro
       environment:
            - CONTAINERS=1

Prerequisites

• Nvidia container toolkit or Nvidia docker 2 installed on the host system
• Nvidia drivers installed on the host system

Container Name Resolution

docker run

• Use the host docker PGID environment variable (999).
• Run grep docker /etc/group | cut -d ':' -f 3 on the host system to get this value.

docker-compose

• Container name resolution no longer requires the host docker PGID and mounting docker.sock. Instead this is handled by HAProxy so that connections are restricted to read-only access. For more information check out the Netdata Docker Installation Page.

Override Directory

This image includes an override feature. Volume mount the container path /etc/netdata anywhere you'd like to store your override files. The override directory contains placeholder directories, a generated netdata.conf file, and the edit-config script. The edit-config script can be used to make edits on any stock conf file. For instance, to edit python.d.conf do the following. docker exec -it netdata /etc/netdata/edit-config python.d.conf This command with load python.d.conf file from the stock configuration directory /usr/lib/netdata/conf.d using vi as an editor. The edited file will be saved to /etc/netdata and will override the stock configuration when netdata is restarted. Subsequent edits on a file will load the file from /etc/netdata.

Notes

• Netdata collects anonymous statistics. If you wish to opt-out, set the envionrment varible DO_NOT_TRACK=1.
• This image uses the default python.d.conf with nvidia_smi: yes uncommented. Volume mount a custom python.d.conf to /etc/netdata/python.d.conf for futher customization.
• If using docker-compose v3+ /etc/docker/daemon.json must edited to make nvidia the default runtime. Example below.
• If using docker-compose v2.4 or previous, add runtime: nvidia.
• Docker v19.03+ has native GPU support built into the runtime. See Docker 19.03 + nvidia-container-toolkit above.

/etc/docker/daemon.json

{
    "default-runtime": "nvidia",
    "runtimes": {
        "nvidia": {
            "path": "/usr/bin/nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}