izenn/docker-intel_gpu_top-telegraf

Intel iGPU utilization to Grafana REWRITE: intel_gpu_top-telegraf

This is a rewrite of intel-gpu-telegraf moving from a fedora base to debian, sending ALL stats from intel_gpu_top to influxdb, changing telegraf to authenticate using influxdb2 tokens, and moving from json to line protocol when sending data through telegraf

This container requires that /dev/dri (the GPU) is piped to the container. You will also need to either run the container with CAP_PERFMON or privileged.

Ultimately the goal was to display GPU statistics via Grafana.

required telegraf.conf modifications:

urls: This is the address to the influxdb server ex: urls = ["http://192.168.1.150:8086"]

token: This is the api token generated from the influxdb dashboard

organization: This is the organization id or name setup in influxdb

bucket: This is the bucket that telegraf will be sending data to

docker-compose.yml modifications:

volumes: this is the path to the scripts and telegraf configuration. this only needs to be changed if you move the intel_gpu_top-telegraf subdirectory

Starting the container

The recommended way to start the container is to build before starting: docker compose build --no-cache then docker compose up -d This is because it will force docker to update the base layer and installed packages in order to squash bugs and security issues

Credits

https://github.com/shubhamcoc/optional_flattening_of_json for the majority of the code needed for flattening the json output from intel_gpu_top