Open ericherdzik-zz opened 6 years ago
Implement a collectd plugin that writes the following metrics to collectd for each GPU device:
collectd
Fan Speed
Temperature
Gpu
Power Readings
Power Draw
Memory Usage
used
total
Utilization
Example implementation using nvidia-smi: https://github.com/bgamari/cuda-collectd Nvidia-smi docs: https://developer.nvidia.com/nvidia-system-management-interface
nvidia-smi
Implement a
collectdplugin that writes the following metrics tocollectdfor each GPU device:Fan Speed(Percent)Temperature.Gpu(Celcius)Power Readings.Power Draw(Watts)Memory Usage.used(MiB)Memory Usage.total(MiB)Utilization.Gpu(Percent)Example implementation using
nvidia-smi: https://github.com/bgamari/cuda-collectd Nvidia-smi docs: https://developer.nvidia.com/nvidia-system-management-interface