mirror of
https://github.com/ClusterCockpit/cc-metric-collector.git
synced 2025-12-15 03:56:15 +01:00
732 B
732 B
nvidia collector
"nvidia": {
"exclude_devices" : [
"0","1"
],
"exclude_metrics": [
"nv_fb_memory",
"nv_fan"
]
}
Metrics:
nv_utilnv_mem_utilnv_mem_totalnv_fb_memorynv_tempnv_fannv_ecc_modenv_perf_statenv_power_usage_reportnv_graphics_clock_reportnv_sm_clock_reportnv_mem_clock_reportnv_max_graphics_clocknv_max_sm_clocknv_max_mem_clocknv_ecc_db_errornv_ecc_sb_errornv_power_man_limitnv_encoder_utilnv_decoder_util
It uses a separate type in the metrics. The output metric looks like this:
<name>,type=accelerator,type-id=<nvidia-gpu-id> value=<metric value> <timestamp>