gpu-monitoring-tools

Open source

Tools for monitoring NVIDIA GPUs on Linux

1.1K stars0 likes0 views

About

Tools for monitoring NVIDIA GPUs on Linux

FINAL UPDATE: August 2021 - This repository has been deprecated.

This repository has been deprecated with the underlying components moved to their own repositories as noted below. The repository has also been archived; setting it to readonly.

The projects have been moved to their own repositories:

This repository will exist for a time to allow for migration to the new repositories.

NVIDIA GPU Monitoring Tools

This repository contains Golang bindings and DCGM-Exporter for gathering GPU telemetry in Kubernetes.

** NOTE: NVML Go bindings have moved to github.com. The NVML Go bindings in this repo are no longer maintained.

** July 2021 - Update #1: The DCGM Go bindings have moved to github.com. The DCGM bindings in this repo are no longer maintained and will eventually be removed.

** June 2021 - NOTICE: Some of the tools in this repository are graduating to their own repos. In the next few weeks both the DCGM Go bindings and the DCGM Exporter will be migrating to github.com/NVIDIA. This will allow for independent versioning, issues, MRs, etc. Efforts will be made to review the existing MRs and issues before the migration occurs.**

Bindings

Golang bindings are provided for the following two libraries:

NVIDIA Management Library (NVML) is a C-based API for monitoring and managing NVIDIA GPU devices.
NVIDIA Data Center GPU Manager (DCGM) is a set of tools for managing and monitoring NVIDIA GPUs in cluster environments. It's a low overhead tool suite that performs a variety of functions on each host system including active health monitoring, diagnostics, system validation, policies, power and clock management, group configuration and accounting.

You will also find samples for both of these bindings in this repository.

DCGM-Exporter

The repository also contains DCGM-Exporter. It exposes GPU metrics exporter for Prometheus leveraging NVIDIA DCGM.

Quickstart

To gather metrics on a GPU node, simply start the dcgm-exporter container:

…

Quickstart on Kubernetes

Note: Consider using the NVIDIA GPU Operator rather than DCGM-Exporter directly.

Ensure you have already setup your cluster with the default runtime as NVIDIA.

The recommended way to install DCGM-Exporter is to use the Helm chart:

bash

$ helm repo add gpu-helm-charts \
  https://nvidia.github.io/gpu-monitoring-tools/helm-charts

Update the repo:

$ helm repo update

And install the chart:

$ helm install \ 
    --generate-name \ 
    gpu-helm-charts/dcgm-exporter

Once the dcgm-exporter pod is deployed, you can use port forwarding to obtain metrics quickly:

…

To integrate DCGM-Exporter with Prometheus and Grafana, see the full instructions in the user guide. dcgm-exporter is deployed as part of the GPU Operator. To get started with integrating with Prometheus, check the Operator user guide.

Building from Source

dcgm-exporter is actually fairly straightforward to build and use. Ensure you have the following:

…

Changing Metrics

With dcgm-exporter you can configure which fields are collected by specifying a custom CSV file. You will find the default CSV file under etc/dcgm-exporter/default-counters.csv in the repository, which is copied on your system or container at /etc/dcgm-exporter/default-counters.csv

The format of this file is pretty straightforward:

# Format,,
# If line starts with a '#' it is considered a comment,,
# DCGM FIELD, Prometheus metric type, help message

# Clocks,,
DCGM_FI_DEV_SM_CLOCK,  gauge, SM clock frequency (in MHz).
DCGM_FI_DEV_MEM_CLOCK, gauge, Memory clock frequency (in MHz).

A custom csv file can be specified using the -f option or --collectors as follows:

$ dcgm-exporter -f /tmp/custom-collectors.csv

Notes:

Always make sure your entries have 3 commas (',')
The complete list of counters that can be collected can be found on the DCGM API reference manual: https://docs.nvidia.com/datacenter/dcgm/latest/dcgm-api/group__dcgmFieldIdentifiers.html