This is a Prometheus Exporter for
exporting NVIDIA GPU metrics. It uses the Go bindings
for NVIDIA Management Library
(NVML) which is a C-based API that can be used for monitoring NVIDIA GPU devices.
Unlike some other similar exporters, it does not call the
nvidia-smi binary.
The repository includes nvml.h, so there are no special requirements from the
build environment. go get should be able to build the exporter binary.
go get github.com/mindprince/nvidia_gpu_prometheus_exporter
The exporter requires the following:
- access to NVML library (
libnvidia-ml.so.1). - access to the GPU devices.
To make sure that the exporter can access the NVML libraries, either add them
to the search path for shared libraries. Or set LD_LIBRARY_PATH to point to
their location.
By default the metrics are exposed on port 9445. This can be updated using
the -web.listen-address flag.