Skip to main content

Nvidia Data Center GPU Manager (DCGM)

Plugin: go.d.plugin Module: dcgm

Overview

This collector gathers NVIDIA GPU telemetry from a dcgm-exporter endpoint. It supports all numeric fields exposed by the exporter and maps them into Netdata-native contexts.

It collects metrics by periodically scraping the exporter Prometheus endpoint over HTTP.

This collector is supported on all platforms.

This collector supports collecting metrics from multiple instances of this integration, including remote instances.

Nvidia Data Center GPU Manager (DCGM) can be monitored further using the following other integrations:

Default Behavior

Auto-Detection

This integration does not support auto-detection in v1.

Limits

The collector applies global and per-metric time series limits to prevent excessive cardinality.

Performance Impact

The impact depends on dcgm-exporter field selection and resulting series cardinality.

Setup

You can configure the dcgm collector in two ways:

MethodBest forHow to
UIFast setup without editing filesGo to Nodes → Configure this node → Collectors → Jobs, search for dcgm, then click + to add a job.
FileIf you prefer configuring via file, or need to automate deployments (e.g., with Ansible)Edit go.d/dcgm.conf and add a job.
important

UI configuration requires paid Netdata Cloud plan.

Prerequisites

Run dcgm-exporter

Install DCGM and run dcgm-exporter so that a Prometheus endpoint is available (default :9400/metrics).

Configure exporter field list

The default exporter profile exposes a small subset of fields. Use the Netdata recommended profile: dcgm-exporter-netdata.csv (raw download: https://raw.githubusercontent.com/netdata/netdata/master/src/go/plugin/go.d/collector/dcgm/dcgm-exporter-netdata.csv).

The Netdata profile enables 127 fields by default and documents all remaining known DCGM fields as commented entries. To customize beyond the baseline, uncomment the field you need and comment one currently enabled field.

Runtime validation artifact: src/go/plugin/go.d/collector/dcgm/runtime-validation-driver-590.48.01-dcgm-exporter-4.4.1-4.5.2.md and src/go/plugin/go.d/collector/dcgm/runtime-validation-driver-590.48.01-dcgm-exporter-4.4.1-4.5.2.json

Validation is primarily version-scoped (NVIDIA driver + DCGM/DCGM-exporter versions), so treat it as a strong baseline rather than universal compatibility.

Example: dcgm-exporter -f /path/to/dcgm-exporter-netdata.csv

Keep collection intervals aligned

Set Netdata update_every to the same value as dcgm-exporter collection interval (default 30 seconds). Example exporter interval: dcgm-exporter -c 30000 and Netdata update_every: 30.

Enable profiling capabilities (optional)

Profiling fields may require additional privileges/capabilities in your runtime environment.

Configuration

Options

The following options can be defined globally: update_every, autodetection_retry.

Config options
GroupOptionDescriptionDefaultRequired
Collectionupdate_everyData collection interval (seconds). Keep this aligned with dcgm-exporter collection interval.30no
autodetection_retryAutodetection retry interval (seconds). Set 0 to disable.0no
TargeturlDCGM exporter metrics endpoint URL.http://127.0.0.1:9400/metricsyes
timeoutHTTP request timeout (seconds).10no
Limitsmax_time_seriesGlobal time series limit. If exceeded, collection is skipped for this cycle.2000no
max_time_series_per_metricPer-metric time series limit. Metrics above this limit are skipped.200no
HTTP AuthusernameUsername for Basic HTTP authentication.no
passwordPassword for Basic HTTP authentication.no
bearer_token_filePath to a file containing a bearer token.no
TLStls_skip_verifySkip TLS certificate and hostname verification (insecure).nono
tls_caPath to CA bundle used to validate the server certificate.no
tls_certPath to client TLS certificate (for mTLS).no
tls_keyPath to client TLS private key (for mTLS).no
Proxyproxy_urlHTTP proxy URL.no
proxy_usernameUsername for proxy authentication.no
proxy_passwordPassword for proxy authentication.no
RequestheadersAdditional HTTP headers to include in the request.no
methodHTTP method.GETno
bodyHTTP request body.no
not_follow_redirectsDo not follow HTTP redirects.nono
force_http2Force HTTP/2 (including h2c over TCP).nono
Virtual NodevnodeAssociate this job with a Virtual Node.no

via UI

Configure the dcgm collector from the Netdata web interface:

  1. Go to Nodes.
  2. Select the node where you want the dcgm data-collection job to run and click the (Configure this node). That node will run the data collection.
  3. The Collectors → Jobs view opens by default.
  4. In the Search box, type dcgm (or scroll the list) to locate the dcgm collector.
  5. Click the + next to the dcgm collector to add a new job.
  6. Fill in the job fields, then click Test to verify the configuration and Submit to save.
    • Test runs the job with the provided settings and shows whether data can be collected.
    • If it fails, an error message appears with details (for example, connection refused, timeout, or command execution errors), so you can adjust and retest.

via File

The configuration file name for this integration is go.d/dcgm.conf.

The file format is YAML. Generally, the structure is:

update_every: 1
autodetection_retry: 0
jobs:
- name: some_name1
- name: some_name2

You can edit the configuration file using the edit-config script from the Netdata config directory.

cd /etc/netdata 2>/dev/null || cd /opt/netdata/etc/netdata
sudo ./edit-config go.d/dcgm.conf
Examples
Local exporter

Collect metrics from a local dcgm-exporter endpoint.

Config
jobs:
- name: local
url: http://127.0.0.1:9400/metrics
update_every: 30

TLS endpoint

Collect metrics over HTTPS with custom CA certificate.

Config
jobs:
- name: secure
url: https://dcgm-exporter.example.com:9400/metrics
update_every: 30
tls_ca: /etc/netdata/certs/dcgm-ca.crt

Increased cardinality limits

Increase limits when collecting large field sets and multiple entities.

Config
jobs:
- name: dcgm_large
url: http://127.0.0.1:9400/metrics
update_every: 30
max_time_series: 10000
max_time_series_per_metric: 2000

Alerts

The following alerts are available:

Alert nameOn metricDescription
dcgm_gpu_xid_errors dcgm.gpu.reliability.xidNVIDIA driver reported GPU XID error on GPU ${label:gpu}
dcgm_gpu_row_remap_failure dcgm.gpu.reliability.row_remap_statusGPU row remapping failed on GPU ${label:gpu}
dcgm_gpu_uncorrectable_remapped_rows dcgm.gpu.reliability.row_remap_eventsUncorrectable remapped rows increased on GPU ${label:gpu}
dcgm_gpu_power_violation dcgm.gpu.throttle.violationsPower throttling detected on GPU ${label:gpu}
dcgm_gpu_thermal_violation dcgm.gpu.throttle.violationsThermal throttling detected on GPU ${label:gpu}

Metrics

Metrics grouped by scope.

The scope defines the instance that the metric belongs to. An instance is uniquely identified by a set of labels.

Metrics are grouped into static Netdata contexts. Contexts are created only when matching DCGM fields are present in the exporter output.

Per gpu

These metrics refer to GPU device instances.

Labels:

LabelDescription
gpugpu label from exporter metrics.
uuiduuid label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.gpu.capability.supportGPU Capability Support metrics.cc_mode, cuda_compute_capability, gpm_support, mig_attributes, mig_ci_info, mig_gi_info, mig_max_slices, supported_clocks, supported_type_infostate
dcgm.gpu.clock.frequencyGPU Clock Frequency metrics.app_mem_clock, app_sm_clock, max_mem_clock, max_sm_clock, max_video_clock, memory, sm, video_clockMHz
dcgm.gpu.compute.activityGPU Compute Pipeline Activity metrics.dram, fp16, fp32, fp64, graphics_engine_active, integer, sm_active, sm_occupancy, tensor%
dcgm.gpu.compute.tensor.activityGPU Tensor Core Activity by precision type.tensor_dfma, tensor_hmma, tensor_imma%
dcgm.gpu.compute.media.activityGPU Media Engine Activity metrics.nvdec0_active, nvdec1_active, nvdec2_active, nvdec3_active, nvdec4_active, nvdec5_active, nvdec6_active, nvdec7_active, nvjpg0_active, nvjpg1_active, nvjpg2_active, nvjpg3_active, nvjpg4_active, nvjpg5_active, nvjpg6_active, nvjpg7_active, nvofa0_active, nvofa1_active%
dcgm.gpu.compute.cache.activityGPU Memory Cache Hit/Miss metrics.hostmem_cache_hit, hostmem_cache_miss, peermem_cache_hit, peermem_cache_missevents/s
dcgm.gpu.compute.utilizationGPU Compute Utilization metrics.decoder, encoder, gpu, memory_copy%
dcgm.gpu.cpu.powerGPU CPU Power metrics.module_power_util_current, sysio_power_util_currentWatts
dcgm.gpu.cpu.infoGPU CPU Information metrics.cpu_model, cpu_vendorvalue
dcgm.gpu.diagnostics.resultsGPU Diagnostics Results metrics.diag_diagnostic_result, diag_eud_result, diag_memory_bandwidth_result, diag_memory_result, diag_memtest_result, diag_nccl_tests_result, diag_nvbandwidth_result, diag_pulse_test_result, diag_software_result, diag_targeted_power_result, diag_targeted_stress_resultstate
dcgm.gpu.diagnostics.statusGPU Diagnostics Status metrics.diag_statusstate
dcgm.gpu.health.statusGPU Health Status metrics.imex_daemon_status, imex_domain_statusstate
dcgm.gpu.interconnect.connectx.error_statusGPU ConnectX Error Status metrics.connectx_correctable_err_mask, connectx_correctable_err_status, connectx_uncorrectable_err_mask, connectx_uncorrectable_err_severity, connectx_uncorrectable_err_statusstate
dcgm.gpu.interconnect.connectx.errorsGPU ConnectX Errors metrics.connectx_correctable_err_mask, connectx_correctable_err_status, connectx_uncorrectable_err_mask, connectx_uncorrectable_err_severity, connectx_uncorrectable_err_statuserrors/s
dcgm.gpu.interconnect.connectx.linkGPU ConnectX Link metrics.connectx_active_pcie_link_speed, connectx_expect_pcie_link_speedvalue
dcgm.gpu.interconnect.connectx.statusGPU ConnectX Status metrics.connectx_healthstate
dcgm.gpu.interconnect.error_rateGPU Interconnect Error Rate metrics.c2c_link_error_intr, c2c_link_error_replay, c2c_link_error_replay_b2berrors/s
dcgm.gpu.interconnect.fabricGPU Fabric State metrics.fabric_clique_id, fabric_cluster_uuid, fabric_health_mask, fabric_manager_error_code, fabric_manager_statusstate
dcgm.gpu.interconnect.nvlink.error_rateGPU NVLink Error Rate metrics.gpu_nvlink_errorserrors/s
dcgm.gpu.interconnect.pcie.error_rateGPU PCIe Error Rate metrics.pcie_count_correctable_errors, pcie_replayerrors/s
dcgm.gpu.interconnect.pcie.link.generationGPU PCIe Link Generation metrics.link_gen, max_link_gengeneration
dcgm.gpu.interconnect.pcie.link.widthGPU PCIe Link Width metrics.connectx_active_pcie_link_width, connectx_expect_pcie_link_width, link_width, max_link_widthlanes
dcgm.gpu.interconnect.stateGPU Interconnect State metrics.c2c_link, c2c_link_power_state, c2c_link_statusstate
dcgm.gpu.interconnect.pcie.stateGPU PCIe State metrics.diag_pcie_resultstate
dcgm.gpu.interconnect.throughputGPU Interconnect Throughput metrics.c2c_max_bandwidth, c2c_rx_all_bytes, c2c_rx_data_bytes, c2c_tx_all_bytes, c2c_tx_data_bytesB/s
dcgm.gpu.interconnect.pcie.throughputGPU PCIe Throughput metrics.pcie_rx, pcie_rx_throughput, pcie_tx, pcie_tx_throughputB/s
dcgm.gpu.interconnect.nvlink.throughputGPU NVLink Throughput metrics.nvlink_rx, nvlink_txB/s
dcgm.gpu.interconnect.total.throughputGPU Interconnect Total Throughput metrics.pcie, nvlinkB/s
dcgm.gpu.internal.boundaryGPU Internal Boundary Fields metrics.first_connectx_field_id, first_vgpu_field_id, internal_fields_0_end, internal_fields_0_start, last_connectx_field_id, last_vgpu_field_idstate
dcgm.gpu.inventory.identityGPU Inventory Identity metrics.brand, count, cuda_visible_devices_str, minor_number, name, nvml_index, serial, uuidvalue
dcgm.gpu.inventory.platformGPU Platform Inventory metrics.platform_chassis_serial_number, platform_chassis_slot_number, platform_host_id, platform_infiniband_guid, platform_module_id, platform_peer_type, platform_tray_indexvalue
dcgm.gpu.inventory.softwareGPU Software and Firmware metrics.inforom_config_check, inforom_config_valid, inforom_image_ver, oem_inforom_ver, power_inforom_ver, process_name, vbios_versionvalue
dcgm.gpu.memory.bar1_usageGPU BAR1 Memory Usage metrics.free, usedB
dcgm.gpu.memory.bar1_capacityGPU BAR1 Memory Capacity metrics.totalB
dcgm.gpu.memory.ecc_error_rateGPU ECC Error Rate metrics.ecc_current, ecc_dbe_agg, ecc_dbe_agg_cbu, ecc_dbe_agg_dev, ecc_dbe_agg_l1, ecc_dbe_agg_l2, ecc_dbe_agg_reg, ecc_dbe_agg_shm, ecc_dbe_agg_srm, ecc_dbe_agg_tex, ecc_dbe_vol, ecc_dbe_vol_cbu, ecc_dbe_vol_dev, ecc_dbe_vol_l1, ecc_dbe_vol_l2, ecc_dbe_vol_reg, ecc_dbe_vol_shm, ecc_dbe_vol_srm, ecc_dbe_vol_tex, ecc_pending, ecc_sbe_agg, ecc_sbe_agg_cbu, ecc_sbe_agg_dev, ecc_sbe_agg_l1, ecc_sbe_agg_l2, ecc_sbe_agg_reg, ecc_sbe_agg_shm, ecc_sbe_agg_srm, ecc_sbe_agg_tex, ecc_sbe_vol, ecc_sbe_vol_cbu, ecc_sbe_vol_dev, ecc_sbe_vol_l1, ecc_sbe_vol_l2, ecc_sbe_vol_reg, ecc_sbe_vol_shm, ecc_sbe_vol_srm, ecc_sbe_vol_texerrors/s
dcgm.gpu.memory.ecc_errorsGPU ECC Errors metrics.ecc_current, ecc_dbe_agg_cbu, ecc_dbe_agg_dev, ecc_dbe_agg_l1, ecc_dbe_agg_l2, ecc_dbe_agg_reg, ecc_dbe_agg_shm, ecc_dbe_agg_srm, ecc_dbe_agg_tex, ecc_dbe_vol_cbu, ecc_dbe_vol_dev, ecc_dbe_vol_l1, ecc_dbe_vol_l2, ecc_dbe_vol_reg, ecc_dbe_vol_shm, ecc_dbe_vol_srm, ecc_dbe_vol_tex, ecc_inforom_ver, ecc_pending, ecc_sbe_agg_cbu, ecc_sbe_agg_dev, ecc_sbe_agg_l1, ecc_sbe_agg_l2, ecc_sbe_agg_reg, ecc_sbe_agg_shm, ecc_sbe_agg_srm, ecc_sbe_agg_tex, ecc_sbe_vol_cbu, ecc_sbe_vol_dev, ecc_sbe_vol_l1, ecc_sbe_vol_l2, ecc_sbe_vol_reg, ecc_sbe_vol_shm, ecc_sbe_vol_srm, ecc_sbe_vol_texerrors
dcgm.gpu.memory.page_retirementsGPU Retired Memory Pages metrics.retired_dbe, retired_pending, retired_sbepages/s
dcgm.gpu.memory.usageGPU Memory Usage metrics.free, reserved, usedB
dcgm.gpu.memory.capacityGPU Memory Capacity metrics.totalB
dcgm.gpu.memory.utilizationGPU Memory Utilization metrics.used_percent%
dcgm.gpu.power.energyGPU Energy Consumption Rate metrics.totalmJ/s
dcgm.gpu.power.profilesGPU Power Profiles metrics.enforced_power_profile_mask, requested_power_profile_mask, valid_power_profile_maskstate
dcgm.gpu.power.smoothingGPU Power Smoothing metrics.pwr_smoothing_active_preset_profile, pwr_smoothing_admin_override_percent_tmp_floor, pwr_smoothing_admin_override_ramp_down_hyst_val, pwr_smoothing_admin_override_ramp_down_rate, pwr_smoothing_admin_override_ramp_up_rate, pwr_smoothing_applied_tmp_ceil, pwr_smoothing_applied_tmp_floor, pwr_smoothing_enabled, pwr_smoothing_hw_circuitry_percent_lifetime_remaining, pwr_smoothing_imm_ramp_down_enabled, pwr_smoothing_max_num_preset_profiles, pwr_smoothing_max_percent_tmp_floor_setting, pwr_smoothing_min_percent_tmp_floor_setting, pwr_smoothing_priv_lvl, pwr_smoothing_profile_percent_tmp_floor, pwr_smoothing_profile_ramp_down_hyst_val, pwr_smoothing_profile_ramp_down_rate, pwr_smoothing_profile_ramp_up_ratevalue
dcgm.gpu.power.usageGPU Power Usage metrics.draw, enforced_limit, power_mgmt_limit, power_mgmt_limit_def, power_mgmt_limit_max, power_mgmt_limit_min, power_usage_instantWatts
dcgm.gpu.reliability.memory_healthGPU Memory Health metrics.banks_remap_rows_avail_high, banks_remap_rows_avail_low, banks_remap_rows_avail_max, banks_remap_rows_avail_none, banks_remap_rows_avail_partial, memory_unrepairable_flag, threshold_srmstate
dcgm.gpu.reliability.recovery_actionGPU Recovery Action metrics.get_gpu_recovery_actionstate
dcgm.gpu.reliability.row_remap_eventsGPU Row Remap Events metrics.correctable_remapped_rows, uncorrectable_remapped_rowsrows/s
dcgm.gpu.reliability.row_remap_statusGPU Row Remap Status metrics.row_remap_failure, row_remap_pendingstate
dcgm.gpu.reliability.xidGPU XID Errors metrics.xidcode
dcgm.gpu.state.configurationGPU Configuration State metrics.autoboost, compute_mode, persistence_mode, sync_boost, sync_boost_violationstate
dcgm.gpu.state.performanceGPU Performance State metrics.pstatestate
dcgm.gpu.state.virtualizationGPU Virtualization State metrics.mig_mode, virtual_modestate
dcgm.gpu.thermal.fan_speedGPU Fan Speed metrics.fan_speed%
dcgm.gpu.thermal.temperatureGPU Temperature metrics.connectx_device_temperature, gpu, gpu_max_op_temp, gpu_temp_limit, mem_max_op_temp, memory, shutdown_temp, slowdown_tempCelsius
dcgm.gpu.throttle.reasonsGPU Throttle Reasons metrics.clocks_event_reasonsbitmask
dcgm.gpu.throttle.violationsGPU Throttle Violation Duration metrics.board_limit_violation, hw_power_brake_slowdown, hw_therm_slowdown, low_utilization_violation, power_violation, reliability_violation, sw_power_cap, sw_therm_slowdown, sync_boost, thermal_violation, total_app_clocks_violation, total_base_clocks_violationmilliseconds/s
dcgm.gpu.topology.affinityGPU Topology and Affinity metrics.cpu_affinity_0, cpu_affinity_1, cpu_affinity_2, cpu_affinity_3, gpu_topology_affinity, gpu_topology_pci, mem_affinity_0, mem_affinity_1, mem_affinity_2, mem_affinity_3, pci_busid, pci_combined_id, pci_subsys_idvalue
dcgm.gpu.virtualization.vgpu.frame_rateGPU vGPU Frame Rate metrics.vgpu_frame_rate_limitfps
dcgm.gpu.virtualization.vgpu.instanceGPU vGPU Instance metrics.vgpu_instance_ids, vgpu_pci_id, vgpu_uuidvalue
dcgm.gpu.virtualization.vgpu.licenseGPU vGPU License metrics.vgpu_instance_license_state, vgpu_license_status, vgpu_type_licensestate
dcgm.gpu.virtualization.vgpu.memoryGPU vGPU Memory metrics.vgpu_memory_usageB
dcgm.gpu.virtualization.vgpu.sessionsGPU vGPU Sessions metrics.vgpu_enc_sessions_info, vgpu_enc_stats, vgpu_fbc_sessions_info, vgpu_fbc_statsvalue
dcgm.gpu.virtualization.vgpu.softwareGPU vGPU Software metrics.vgpu_driver_versionvalue
dcgm.gpu.virtualization.vgpu.typeGPU vGPU Type metrics.creatable_vgpu_type_ids, supported_vgpu_type_ids, vgpu_type, vgpu_type_class, vgpu_type_info, vgpu_type_namevalue
dcgm.gpu.virtualization.vgpu.utilizationGPU vGPU Utilization metrics.vgpu_per_process_utilization%
dcgm.gpu.virtualization.vgpu.vmGPU vGPU VM metrics.vgpu_vm_gpu_instance_id, vgpu_vm_id, vgpu_vm_namevalue
dcgm.gpu.workload.sessionsGPU Workload Sessions metrics.accounting_data, enc_stats, fbc_sessions_info, fbc_statsvalue

Per mig

These metrics refer to MIG instances.

Labels:

LabelDescription
gpugpu label from exporter metrics.
gpu_i_idgpu_i_id label from exporter metrics.
gpu_i_profilegpu_i_profile label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.mig.clock.frequencyMIG Clock Frequency metrics.app_mem_clock, app_sm_clock, max_mem_clock, max_sm_clock, max_video_clock, memory, sm, video_clockMHz
dcgm.mig.compute.activityMIG Compute Pipeline Activity metrics.dram, fp16, fp32, fp64, graphics_engine_active, integer, sm_active, sm_occupancy, tensor%
dcgm.mig.compute.tensor.activityMIG Tensor Core Activity by precision type.tensor_dfma, tensor_hmma, tensor_imma%
dcgm.mig.compute.media.activityMIG Media Engine Activity metrics.nvdec0_active, nvdec1_active, nvdec2_active, nvdec3_active, nvdec4_active, nvdec5_active, nvdec6_active, nvdec7_active, nvjpg0_active, nvjpg1_active, nvjpg2_active, nvjpg3_active, nvjpg4_active, nvjpg5_active, nvjpg6_active, nvjpg7_active, nvofa0_active, nvofa1_active%
dcgm.mig.compute.cache.activityMIG Memory Cache Hit/Miss metrics.hostmem_cache_hit, hostmem_cache_miss, peermem_cache_hit, peermem_cache_missevents/s
dcgm.mig.compute.utilizationMIG Compute Utilization metrics.decoder, encoder, gpu, memory_copy%
dcgm.mig.interconnect.nvlink.berMIG NVLink Bit Error Rate metrics.nvlink_count_effective_ber, nvlink_count_effective_ber_float, nvlink_count_symbol_ber, nvlink_count_symbol_ber_floatratio
dcgm.mig.interconnect.nvlink.congestionMIG NVLink Congestion metrics.nvlink_ppcnt_ibpc_port_xmit_waitevents/s
dcgm.mig.interconnect.error_rateMIG Interconnect Error Rate metrics.c2c_link_error_intr, c2c_link_error_replay, c2c_link_error_replay_b2berrors/s
dcgm.mig.interconnect.nvlink.error_rateMIG NVLink Error Rate metrics.gpu_nvlink_errors, nvlink_count_effective_errors, nvlink_count_fec_history_0, nvlink_count_fec_history_1, nvlink_count_fec_history_10, nvlink_count_fec_history_11, nvlink_count_fec_history_12, nvlink_count_fec_history_13, nvlink_count_fec_history_14, nvlink_count_fec_history_15, nvlink_count_fec_history_2, nvlink_count_fec_history_3, nvlink_count_fec_history_4, nvlink_count_fec_history_5, nvlink_count_fec_history_6, nvlink_count_fec_history_7, nvlink_count_fec_history_8, nvlink_count_fec_history_9, nvlink_count_link_recovery_events, nvlink_count_link_recovery_failed_events, nvlink_count_link_recovery_successful_events, nvlink_count_local_link_integrity_errors, nvlink_count_rx_buffer_overrun_errors, nvlink_count_rx_errors, nvlink_count_rx_general_errors, nvlink_count_rx_malformed_packet_errors, nvlink_count_rx_remote_errors, nvlink_count_rx_symbol_errors, nvlink_count_tx_discards, nvlink_crc_data_error, nvlink_crc_data_error_count_l0, nvlink_crc_data_error_count_l1, nvlink_crc_data_error_count_l10, nvlink_crc_data_error_count_l11, nvlink_crc_data_error_count_l12, nvlink_crc_data_error_count_l13, nvlink_crc_data_error_count_l14, nvlink_crc_data_error_count_l15, nvlink_crc_data_error_count_l16, nvlink_crc_data_error_count_l17, nvlink_crc_data_error_count_l2, nvlink_crc_data_error_count_l3, nvlink_crc_data_error_count_l4, nvlink_crc_data_error_count_l5, nvlink_crc_data_error_count_l6, nvlink_crc_data_error_count_l7, nvlink_crc_data_error_count_l8, nvlink_crc_data_error_count_l9, nvlink_crc_flit_error, nvlink_crc_flit_error_count_l0, nvlink_crc_flit_error_count_l1, nvlink_crc_flit_error_count_l10, nvlink_crc_flit_error_count_l11, nvlink_crc_flit_error_count_l12, nvlink_crc_flit_error_count_l13, nvlink_crc_flit_error_count_l14, nvlink_crc_flit_error_count_l15, nvlink_crc_flit_error_count_l16, nvlink_crc_flit_error_count_l17, nvlink_crc_flit_error_count_l2, nvlink_crc_flit_error_count_l3, nvlink_crc_flit_error_count_l4, nvlink_crc_flit_error_count_l5, nvlink_crc_flit_error_count_l6, nvlink_crc_flit_error_count_l7, nvlink_crc_flit_error_count_l8, nvlink_crc_flit_error_count_l9, nvlink_error_dl_crc, nvlink_error_dl_recovery, nvlink_error_dl_replay, nvlink_ppcnt_physical_successful_recovery_events, nvlink_ppcnt_plr_rcv_uncorrectable_code, nvlink_ppcnt_recovery_time_since_last, nvlink_ppcnt_recovery_total_successful_events, nvlink_pprm_oper_recovery, nvlink_recovery_error, nvlink_recovery_error_count_l0, nvlink_recovery_error_count_l1, nvlink_recovery_error_count_l10, nvlink_recovery_error_count_l11, nvlink_recovery_error_count_l12, nvlink_recovery_error_count_l13, nvlink_recovery_error_count_l14, nvlink_recovery_error_count_l15, nvlink_recovery_error_count_l16, nvlink_recovery_error_count_l17, nvlink_recovery_error_count_l2, nvlink_recovery_error_count_l3, nvlink_recovery_error_count_l4, nvlink_recovery_error_count_l5, nvlink_recovery_error_count_l6, nvlink_recovery_error_count_l7, nvlink_recovery_error_count_l8, nvlink_recovery_error_count_l9, nvlink_replay_error, nvlink_replay_error_count_l0, nvlink_replay_error_count_l1, nvlink_replay_error_count_l10, nvlink_replay_error_count_l11, nvlink_replay_error_count_l12, nvlink_replay_error_count_l13, nvlink_replay_error_count_l14, nvlink_replay_error_count_l15, nvlink_replay_error_count_l16, nvlink_replay_error_count_l17, nvlink_replay_error_count_l2, nvlink_replay_error_count_l3, nvlink_replay_error_count_l4, nvlink_replay_error_count_l5, nvlink_replay_error_count_l6, nvlink_replay_error_count_l7, nvlink_replay_error_count_l8, nvlink_replay_error_count_l9errors/s
dcgm.mig.interconnect.pcie.error_rateMIG PCIe Error Rate metrics.pcie_count_correctable_errors, pcie_replayerrors/s
dcgm.mig.interconnect.nvlink.errorsMIG NVLink Errors metrics.nvlink_ppcnt_plr_rcv_uncorrectable_codeerrors
dcgm.mig.interconnect.fabricMIG Fabric State metrics.fabric_clique_id, fabric_cluster_uuid, fabric_health_mask, fabric_manager_error_code, fabric_manager_statusstate
dcgm.mig.interconnect.pcie.link.generationMIG PCIe Link Generation metrics.link_gen, max_link_gengeneration
dcgm.mig.interconnect.pcie.link.widthMIG PCIe Link Width metrics.link_width, max_link_widthlanes
dcgm.mig.interconnect.stateMIG Interconnect State metrics.c2c_link, c2c_link_power_state, c2c_link_statusstate
dcgm.mig.interconnect.pcie.stateMIG PCIe State metrics.diag_pcie_resultstate
dcgm.mig.interconnect.nvlink.stateMIG NVLink State metrics.gpu_topology_nvlink, nvlink_get_state, nvlink_ppcnt_physical_link_down_counter, nvlink_ppcnt_plr_rcv_code_err, nvlink_ppcnt_plr_sync_events, nvlink_ppcnt_plr_xmit_retry_events, p2p_nvlink_statusstate
dcgm.mig.interconnect.throughputMIG Interconnect Throughput metrics.c2c_max_bandwidth, c2c_rx_all_bytes, c2c_rx_data_bytes, c2c_tx_all_bytes, c2c_tx_data_bytesB/s
dcgm.mig.interconnect.nvlink.throughputMIG NVLink Throughput metrics.nvlink_bandwidth_l0, nvlink_bandwidth_l1, nvlink_bandwidth_l10, nvlink_bandwidth_l11, nvlink_bandwidth_l12, nvlink_bandwidth_l13, nvlink_bandwidth_l14, nvlink_bandwidth_l15, nvlink_bandwidth_l16, nvlink_bandwidth_l17, nvlink_bandwidth_l2, nvlink_bandwidth_l3, nvlink_bandwidth_l4, nvlink_bandwidth_l5, nvlink_bandwidth_l6, nvlink_bandwidth_l7, nvlink_bandwidth_l8, nvlink_bandwidth_l9, nvlink_count_rx, nvlink_count_tx, nvlink_l0_rx, nvlink_l0_tx, nvlink_l10_rx, nvlink_l10_tx, nvlink_l11_rx, nvlink_l11_tx, nvlink_l12_rx, nvlink_l12_tx, nvlink_l13_rx, nvlink_l13_tx, nvlink_l14_rx, nvlink_l14_tx, nvlink_l15_rx, nvlink_l15_tx, nvlink_l16_rx, nvlink_l16_tx, nvlink_l17_rx, nvlink_l17_tx, nvlink_l1_rx, nvlink_l1_tx, nvlink_l2_rx, nvlink_l2_tx, nvlink_l3_rx, nvlink_l3_tx, nvlink_l4_rx, nvlink_l4_tx, nvlink_l5_rx, nvlink_l5_tx, nvlink_l6_rx, nvlink_l6_tx, nvlink_l7_rx, nvlink_l7_tx, nvlink_l8_rx, nvlink_l8_tx, nvlink_l9_rx, nvlink_l9_tx, nvlink_rx_bandwidth, nvlink_rx_bandwidth_l0, nvlink_rx_bandwidth_l1, nvlink_rx_bandwidth_l10, nvlink_rx_bandwidth_l11, nvlink_rx_bandwidth_l12, nvlink_rx_bandwidth_l13, nvlink_rx_bandwidth_l14, nvlink_rx_bandwidth_l15, nvlink_rx_bandwidth_l16, nvlink_rx_bandwidth_l17, nvlink_rx_bandwidth_l2, nvlink_rx_bandwidth_l3, nvlink_rx_bandwidth_l4, nvlink_rx_bandwidth_l5, nvlink_rx_bandwidth_l6, nvlink_rx_bandwidth_l7, nvlink_rx_bandwidth_l8, nvlink_rx_bandwidth_l9, nvlink_rx, nvlink_tx_bandwidth, nvlink_tx_bandwidth_l0, nvlink_tx_bandwidth_l1, nvlink_tx_bandwidth_l10, nvlink_tx_bandwidth_l11, nvlink_tx_bandwidth_l12, nvlink_tx_bandwidth_l13, nvlink_tx_bandwidth_l14, nvlink_tx_bandwidth_l15, nvlink_tx_bandwidth_l16, nvlink_tx_bandwidth_l17, nvlink_tx_bandwidth_l2, nvlink_tx_bandwidth_l3, nvlink_tx_bandwidth_l4, nvlink_tx_bandwidth_l5, nvlink_tx_bandwidth_l6, nvlink_tx_bandwidth_l7, nvlink_tx_bandwidth_l8, nvlink_tx_bandwidth_l9, nvlink_txB/s
dcgm.mig.interconnect.pcie.throughputMIG PCIe Throughput metrics.pcie_rx, pcie_rx_throughput, pcie_tx, pcie_tx_throughputB/s
dcgm.mig.interconnect.total.throughputMIG Interconnect Total Throughput metrics.pcie, nvlinkB/s
dcgm.mig.interconnect.nvlink.trafficMIG NVLink Traffic metrics.nvlink_count_rx_packets, nvlink_count_tx_packets, nvlink_ppcnt_plr_rcv_codes, nvlink_ppcnt_plr_xmit_codes, nvlink_ppcnt_plr_xmit_retry_codesevents/s
dcgm.mig.memory.bar1_usageMIG BAR1 Memory Usage metrics.free, usedB
dcgm.mig.memory.bar1_capacityMIG BAR1 Memory Capacity metrics.totalB
dcgm.mig.memory.ecc_error_rateMIG ECC Error Rate metrics.ecc_current, ecc_dbe_agg, ecc_dbe_agg_cbu, ecc_dbe_agg_dev, ecc_dbe_agg_l1, ecc_dbe_agg_l2, ecc_dbe_agg_reg, ecc_dbe_agg_shm, ecc_dbe_agg_srm, ecc_dbe_agg_tex, ecc_dbe_vol, ecc_dbe_vol_cbu, ecc_dbe_vol_dev, ecc_dbe_vol_l1, ecc_dbe_vol_l2, ecc_dbe_vol_reg, ecc_dbe_vol_shm, ecc_dbe_vol_srm, ecc_dbe_vol_tex, ecc_pending, ecc_sbe_agg, ecc_sbe_agg_cbu, ecc_sbe_agg_dev, ecc_sbe_agg_l1, ecc_sbe_agg_l2, ecc_sbe_agg_reg, ecc_sbe_agg_shm, ecc_sbe_agg_srm, ecc_sbe_agg_tex, ecc_sbe_vol, ecc_sbe_vol_cbu, ecc_sbe_vol_dev, ecc_sbe_vol_l1, ecc_sbe_vol_l2, ecc_sbe_vol_reg, ecc_sbe_vol_shm, ecc_sbe_vol_srm, ecc_sbe_vol_tex, nvlink_ecc_data_errorerrors/s
dcgm.mig.memory.ecc_errorsMIG ECC Errors metrics.ecc_current, ecc_dbe_agg_cbu, ecc_dbe_agg_dev, ecc_dbe_agg_l1, ecc_dbe_agg_l2, ecc_dbe_agg_reg, ecc_dbe_agg_shm, ecc_dbe_agg_srm, ecc_dbe_agg_tex, ecc_dbe_vol_cbu, ecc_dbe_vol_dev, ecc_dbe_vol_l1, ecc_dbe_vol_l2, ecc_dbe_vol_reg, ecc_dbe_vol_shm, ecc_dbe_vol_srm, ecc_dbe_vol_tex, ecc_inforom_ver, ecc_pending, ecc_sbe_agg_cbu, ecc_sbe_agg_dev, ecc_sbe_agg_l1, ecc_sbe_agg_l2, ecc_sbe_agg_reg, ecc_sbe_agg_shm, ecc_sbe_agg_srm, ecc_sbe_agg_tex, ecc_sbe_vol_cbu, ecc_sbe_vol_dev, ecc_sbe_vol_l1, ecc_sbe_vol_l2, ecc_sbe_vol_reg, ecc_sbe_vol_shm, ecc_sbe_vol_srm, ecc_sbe_vol_texerrors
dcgm.mig.memory.page_retirementsMIG Retired Memory Pages metrics.retired_dbe, retired_pending, retired_sbepages/s
dcgm.mig.memory.usageMIG Memory Usage metrics.free, reserved, usedB
dcgm.mig.memory.capacityMIG Memory Capacity metrics.totalB
dcgm.mig.memory.utilizationMIG Memory Utilization metrics.used_percent%
dcgm.mig.power.energyMIG Energy Consumption Rate metrics.totalmJ/s
dcgm.mig.power.profilesMIG Power Profiles metrics.enforced_power_profile_mask, requested_power_profile_mask, valid_power_profile_maskstate
dcgm.mig.power.smoothingMIG Power Smoothing metrics.pwr_smoothing_active_preset_profile, pwr_smoothing_admin_override_percent_tmp_floor, pwr_smoothing_admin_override_ramp_down_hyst_val, pwr_smoothing_admin_override_ramp_down_rate, pwr_smoothing_admin_override_ramp_up_rate, pwr_smoothing_applied_tmp_ceil, pwr_smoothing_applied_tmp_floor, pwr_smoothing_enabled, pwr_smoothing_hw_circuitry_percent_lifetime_remaining, pwr_smoothing_imm_ramp_down_enabled, pwr_smoothing_max_num_preset_profiles, pwr_smoothing_max_percent_tmp_floor_setting, pwr_smoothing_min_percent_tmp_floor_setting, pwr_smoothing_priv_lvl, pwr_smoothing_profile_percent_tmp_floor, pwr_smoothing_profile_ramp_down_hyst_val, pwr_smoothing_profile_ramp_down_rate, pwr_smoothing_profile_ramp_up_ratevalue
dcgm.mig.power.usageMIG Power Usage metrics.draw, enforced_limit, power_mgmt_limit, power_mgmt_limit_def, power_mgmt_limit_max, power_mgmt_limit_min, power_usage_instantWatts
dcgm.mig.reliability.memory_healthMIG Memory Health metrics.banks_remap_rows_avail_high, banks_remap_rows_avail_low, banks_remap_rows_avail_max, banks_remap_rows_avail_none, banks_remap_rows_avail_partial, memory_unrepairable_flag, threshold_srmstate
dcgm.mig.reliability.recovery_actionMIG Recovery Action metrics.get_gpu_recovery_actionstate
dcgm.mig.reliability.row_remap_eventsMIG Row Remap Events metrics.correctable_remapped_rows, uncorrectable_remapped_rowsrows/s
dcgm.mig.reliability.row_remap_statusMIG Row Remap Status metrics.row_remap_failure, row_remap_pendingstate
dcgm.mig.reliability.xidMIG XID Errors metrics.xidcode
dcgm.mig.state.configurationMIG Configuration State metrics.autoboost, compute_mode, persistence_mode, sync_boost, sync_boost_violationstate
dcgm.mig.state.performanceMIG Performance State metrics.pstatestate
dcgm.mig.state.virtualizationMIG Virtualization State metrics.mig_mode, virtual_modestate
dcgm.mig.thermal.fan_speedMIG Fan Speed metrics.fan_speed%
dcgm.mig.thermal.temperatureMIG Temperature metrics.gpu, gpu_max_op_temp, gpu_temp_limit, mem_max_op_temp, memory, shutdown_temp, slowdown_tempCelsius
dcgm.mig.throttle.reasonsMIG Throttle Reasons metrics.clocks_event_reasonsbitmask
dcgm.mig.throttle.violationsMIG Throttle Violation Duration metrics.board_limit_violation, hw_power_brake_slowdown, hw_therm_slowdown, low_utilization_violation, power_violation, reliability_violation, sw_power_cap, sw_therm_slowdown, sync_boost, thermal_violation, total_app_clocks_violation, total_base_clocks_violationmilliseconds/s

These metrics refer to NVLink link instances.

Labels:

LabelDescription
gpugpu label from exporter metrics.
gpu_uuidgpu_uuid label from exporter metrics.
nvlinknvlink label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.nvlink.interconnect.berNVLink Interconnect Bit Error Rate metrics.nvlink_count_effective_ber, nvlink_count_effective_ber_float, nvlink_count_symbol_ber, nvlink_count_symbol_ber_floatratio
dcgm.nvlink.interconnect.congestionNVLink Interconnect Congestion metrics.nvlink_ppcnt_ibpc_port_xmit_waitevents/s
dcgm.nvlink.interconnect.error_rateNVLink Interconnect Error Rate metrics.gpu_nvlink_errors, nvlink_count_effective_errors, nvlink_count_fec_history_0, nvlink_count_fec_history_1, nvlink_count_fec_history_10, nvlink_count_fec_history_11, nvlink_count_fec_history_12, nvlink_count_fec_history_13, nvlink_count_fec_history_14, nvlink_count_fec_history_15, nvlink_count_fec_history_2, nvlink_count_fec_history_3, nvlink_count_fec_history_4, nvlink_count_fec_history_5, nvlink_count_fec_history_6, nvlink_count_fec_history_7, nvlink_count_fec_history_8, nvlink_count_fec_history_9, nvlink_count_link_recovery_events, nvlink_count_link_recovery_failed_events, nvlink_count_link_recovery_successful_events, nvlink_count_local_link_integrity_errors, nvlink_count_rx_buffer_overrun_errors, nvlink_count_rx_errors, nvlink_count_rx_general_errors, nvlink_count_rx_malformed_packet_errors, nvlink_count_rx_remote_errors, nvlink_count_rx_symbol_errors, nvlink_count_tx_discards, nvlink_crc_data_error, nvlink_crc_data_error_count_l0, nvlink_crc_data_error_count_l1, nvlink_crc_data_error_count_l10, nvlink_crc_data_error_count_l11, nvlink_crc_data_error_count_l12, nvlink_crc_data_error_count_l13, nvlink_crc_data_error_count_l14, nvlink_crc_data_error_count_l15, nvlink_crc_data_error_count_l16, nvlink_crc_data_error_count_l17, nvlink_crc_data_error_count_l2, nvlink_crc_data_error_count_l3, nvlink_crc_data_error_count_l4, nvlink_crc_data_error_count_l5, nvlink_crc_data_error_count_l6, nvlink_crc_data_error_count_l7, nvlink_crc_data_error_count_l8, nvlink_crc_data_error_count_l9, nvlink_crc_flit_error, nvlink_crc_flit_error_count_l0, nvlink_crc_flit_error_count_l1, nvlink_crc_flit_error_count_l10, nvlink_crc_flit_error_count_l11, nvlink_crc_flit_error_count_l12, nvlink_crc_flit_error_count_l13, nvlink_crc_flit_error_count_l14, nvlink_crc_flit_error_count_l15, nvlink_crc_flit_error_count_l16, nvlink_crc_flit_error_count_l17, nvlink_crc_flit_error_count_l2, nvlink_crc_flit_error_count_l3, nvlink_crc_flit_error_count_l4, nvlink_crc_flit_error_count_l5, nvlink_crc_flit_error_count_l6, nvlink_crc_flit_error_count_l7, nvlink_crc_flit_error_count_l8, nvlink_crc_flit_error_count_l9, nvlink_error_dl_crc, nvlink_error_dl_recovery, nvlink_error_dl_replay, nvlink_ppcnt_physical_successful_recovery_events, nvlink_ppcnt_plr_rcv_uncorrectable_code, nvlink_ppcnt_recovery_time_since_last, nvlink_ppcnt_recovery_total_successful_events, nvlink_pprm_oper_recovery, nvlink_recovery_error, nvlink_recovery_error_count_l0, nvlink_recovery_error_count_l1, nvlink_recovery_error_count_l10, nvlink_recovery_error_count_l11, nvlink_recovery_error_count_l12, nvlink_recovery_error_count_l13, nvlink_recovery_error_count_l14, nvlink_recovery_error_count_l15, nvlink_recovery_error_count_l16, nvlink_recovery_error_count_l17, nvlink_recovery_error_count_l2, nvlink_recovery_error_count_l3, nvlink_recovery_error_count_l4, nvlink_recovery_error_count_l5, nvlink_recovery_error_count_l6, nvlink_recovery_error_count_l7, nvlink_recovery_error_count_l8, nvlink_recovery_error_count_l9, nvlink_replay_error, nvlink_replay_error_count_l0, nvlink_replay_error_count_l1, nvlink_replay_error_count_l10, nvlink_replay_error_count_l11, nvlink_replay_error_count_l12, nvlink_replay_error_count_l13, nvlink_replay_error_count_l14, nvlink_replay_error_count_l15, nvlink_replay_error_count_l16, nvlink_replay_error_count_l17, nvlink_replay_error_count_l2, nvlink_replay_error_count_l3, nvlink_replay_error_count_l4, nvlink_replay_error_count_l5, nvlink_replay_error_count_l6, nvlink_replay_error_count_l7, nvlink_replay_error_count_l8, nvlink_replay_error_count_l9errors/s
dcgm.nvlink.interconnect.errorsNVLink Interconnect Errors metrics.nvlink_ppcnt_plr_rcv_uncorrectable_codeerrors
dcgm.nvlink.interconnect.stateNVLink Interconnect State metrics.gpu_topology_nvlink, nvlink_get_state, nvlink_ppcnt_physical_link_down_counter, nvlink_ppcnt_plr_rcv_code_err, nvlink_ppcnt_plr_sync_events, nvlink_ppcnt_plr_xmit_retry_events, p2p_nvlink_statusstate
dcgm.nvlink.interconnect.throughputNVLink Interconnect Throughput metrics.nvlink_bandwidth, nvlink_bandwidth_l0, nvlink_bandwidth_l1, nvlink_bandwidth_l10, nvlink_bandwidth_l11, nvlink_bandwidth_l12, nvlink_bandwidth_l13, nvlink_bandwidth_l14, nvlink_bandwidth_l15, nvlink_bandwidth_l16, nvlink_bandwidth_l17, nvlink_bandwidth_l2, nvlink_bandwidth_l3, nvlink_bandwidth_l4, nvlink_bandwidth_l5, nvlink_bandwidth_l6, nvlink_bandwidth_l7, nvlink_bandwidth_l8, nvlink_bandwidth_l9, nvlink_count_rx, nvlink_count_tx, nvlink_l0_rx, nvlink_l0_tx, nvlink_l10_rx, nvlink_l10_tx, nvlink_l11_rx, nvlink_l11_tx, nvlink_l12_rx, nvlink_l12_tx, nvlink_l13_rx, nvlink_l13_tx, nvlink_l14_rx, nvlink_l14_tx, nvlink_l15_rx, nvlink_l15_tx, nvlink_l16_rx, nvlink_l16_tx, nvlink_l17_rx, nvlink_l17_tx, nvlink_l1_rx, nvlink_l1_tx, nvlink_l2_rx, nvlink_l2_tx, nvlink_l3_rx, nvlink_l3_tx, nvlink_l4_rx, nvlink_l4_tx, nvlink_l5_rx, nvlink_l5_tx, nvlink_l6_rx, nvlink_l6_tx, nvlink_l7_rx, nvlink_l7_tx, nvlink_l8_rx, nvlink_l8_tx, nvlink_l9_rx, nvlink_l9_tx, nvlink_rx_bandwidth, nvlink_rx_bandwidth_l0, nvlink_rx_bandwidth_l1, nvlink_rx_bandwidth_l10, nvlink_rx_bandwidth_l11, nvlink_rx_bandwidth_l12, nvlink_rx_bandwidth_l13, nvlink_rx_bandwidth_l14, nvlink_rx_bandwidth_l15, nvlink_rx_bandwidth_l16, nvlink_rx_bandwidth_l17, nvlink_rx_bandwidth_l2, nvlink_rx_bandwidth_l3, nvlink_rx_bandwidth_l4, nvlink_rx_bandwidth_l5, nvlink_rx_bandwidth_l6, nvlink_rx_bandwidth_l7, nvlink_rx_bandwidth_l8, nvlink_rx_bandwidth_l9, nvlink_rx, nvlink_tx_bandwidth, nvlink_tx_bandwidth_l0, nvlink_tx_bandwidth_l1, nvlink_tx_bandwidth_l10, nvlink_tx_bandwidth_l11, nvlink_tx_bandwidth_l12, nvlink_tx_bandwidth_l13, nvlink_tx_bandwidth_l14, nvlink_tx_bandwidth_l15, nvlink_tx_bandwidth_l16, nvlink_tx_bandwidth_l17, nvlink_tx_bandwidth_l2, nvlink_tx_bandwidth_l3, nvlink_tx_bandwidth_l4, nvlink_tx_bandwidth_l5, nvlink_tx_bandwidth_l6, nvlink_tx_bandwidth_l7, nvlink_tx_bandwidth_l8, nvlink_tx_bandwidth_l9, nvlink_txB/s
dcgm.nvlink.interconnect.trafficNVLink Interconnect Traffic metrics.nvlink_count_rx_packets, nvlink_count_tx_packets, nvlink_ppcnt_plr_rcv_codes, nvlink_ppcnt_plr_xmit_codes, nvlink_ppcnt_plr_xmit_retry_codesevents/s
dcgm.nvlink.internal.boundaryNVLink Internal Boundary Fields metrics.nvlink_ppcnt_recovery_time_between_last_twostate
dcgm.nvlink.memory.ecc_error_rateNVLink ECC Error Rate metrics.nvlink_ecc_data_errorerrors/s

Per nvswitch

These metrics refer to NVSwitch instances.

Labels:

LabelDescription
nvswitchnvswitch label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.nvswitch.interconnect.nvswitch.currentNVSwitch Current metrics.nvswitch_current_iddq, nvswitch_current_iddq_dvdd, nvswitch_current_iddq_revvalue
dcgm.nvswitch.interconnect.nvswitch.errorsNVSwitch NVSwitch Errors metrics.nvswitch_fatal_errors, nvswitch_link_crc_errors, nvswitch_link_crc_errors_lane0, nvswitch_link_crc_errors_lane1, nvswitch_link_crc_errors_lane2, nvswitch_link_crc_errors_lane3, nvswitch_link_crc_errors_lane4, nvswitch_link_crc_errors_lane5, nvswitch_link_crc_errors_lane6, nvswitch_link_crc_errors_lane7, nvswitch_link_fatal_errors, nvswitch_link_flit_errors, nvswitch_link_non_fatal_errors, nvswitch_link_recovery_errors, nvswitch_link_replay_errors, nvswitch_non_fatal_errorserrors/s
dcgm.nvswitch.interconnect.nvswitch.latencyNVSwitch NVSwitch Link Latency metrics.nvswitch_link_latency_count_vc0, nvswitch_link_latency_count_vc1, nvswitch_link_latency_count_vc2, nvswitch_link_latency_count_vc3, nvswitch_link_latency_high_vc0, nvswitch_link_latency_high_vc1, nvswitch_link_latency_high_vc2, nvswitch_link_latency_high_vc3, nvswitch_link_latency_low_vc0, nvswitch_link_latency_low_vc1, nvswitch_link_latency_low_vc2, nvswitch_link_latency_low_vc3, nvswitch_link_latency_medium_vc0, nvswitch_link_latency_medium_vc1, nvswitch_link_latency_medium_vc2, nvswitch_link_latency_medium_vc3, nvswitch_link_latency_panic_vc0, nvswitch_link_latency_panic_vc1, nvswitch_link_latency_panic_vc2, nvswitch_link_latency_panic_vc3events/s
dcgm.nvswitch.interconnect.nvswitch.powerNVSwitch NVSwitch Power metrics.nvswitch_power_dvdd, nvswitch_power_hvdd, nvswitch_power_vddWatts
dcgm.nvswitch.interconnect.nvswitch.statusNVSwitch NVSwitch Status metrics.nvswitch_link_status, nvswitch_link_type, nvswitch_reset_requiredstate
dcgm.nvswitch.interconnect.nvswitch.throughputNVSwitch NVSwitch Throughput metrics.nvswitch_link_throughput_rx, nvswitch_link_throughput_tx, nvswitch_throughput_rx, nvswitch_throughput_txB/s
dcgm.nvswitch.interconnect.nvswitch.topologyNVSwitch NVSwitch Topology metrics.nvswitch_device_uuid, nvswitch_link_device_link_id, nvswitch_link_device_link_sid, nvswitch_link_id, nvswitch_link_remote_pcie_bus, nvswitch_link_remote_pcie_device, nvswitch_link_remote_pcie_domain, nvswitch_link_remote_pcie_function, nvswitch_pcie_bus, nvswitch_pcie_device, nvswitch_pcie_domain, nvswitch_pcie_function, nvswitch_phys_idvalue
dcgm.nvswitch.interconnect.nvswitch.voltageNVSwitch NVSwitch Voltage metrics.nvswitch_voltage_mvoltmV
dcgm.nvswitch.internal.boundaryNVSwitch Internal Boundary Fields metrics.first_nvswitch_field_id, last_nvswitch_field_idstate
dcgm.nvswitch.memory.ecc_error_rateNVSwitch ECC Error Rate metrics.nvswitch_link_ecc_errors, nvswitch_link_ecc_errors_lane0, nvswitch_link_ecc_errors_lane1, nvswitch_link_ecc_errors_lane2, nvswitch_link_ecc_errors_lane3, nvswitch_link_ecc_errors_lane4, nvswitch_link_ecc_errors_lane5, nvswitch_link_ecc_errors_lane6, nvswitch_link_ecc_errors_lane7errors/s
dcgm.nvswitch.thermal.temperatureNVSwitch Temperature metrics.nvswitch_temperature_current, nvswitch_temperature_limit_shutdown, nvswitch_temperature_limit_slowdownCelsius

Per cpu

These metrics refer to host CPU instances.

Labels:

LabelDescription
cpucpu label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.cpu.clock.frequencyCPU Clock Frequency metrics.cpu_clock_currentMHz
dcgm.cpu.cpu.infoCPU CPU Information metrics.cpu_model, cpu_vendorvalue
dcgm.cpu.cpu.powerCPU CPU Power metrics.cpu_power_limit, cpu_power_util_currentWatts
dcgm.cpu.cpu.temperatureCPU CPU Temperature metrics.cpu_temp_critical, cpu_temp_current, cpu_temp_warningCelsius
dcgm.cpu.cpu.utilizationCPU CPU Utilization metrics.cpu_util, cpu_util_irq, cpu_util_nice, cpu_util_sys, cpu_util_user%
dcgm.cpu.diagnostics.resultsCPU Diagnostics Results metrics.diag_cpu_eud_resultstate

Per cpu_core

These metrics refer to host CPU core instances.

Labels:

LabelDescription
cpucpu label from exporter metrics.
cpucorecpucore label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.cpu_core.clock.frequencyCPU Core Clock Frequency metrics.cpu_clock_currentMHz
dcgm.cpu_core.cpu.infoCPU Core CPU Information metrics.cpu_model, cpu_vendorvalue
dcgm.cpu_core.cpu.powerCPU Core CPU Power metrics.cpu_power_limit, cpu_power_util_currentWatts
dcgm.cpu_core.cpu.temperatureCPU Core CPU Temperature metrics.cpu_temp_critical, cpu_temp_current, cpu_temp_warningCelsius
dcgm.cpu_core.cpu.utilizationCPU Core CPU Utilization metrics.cpu_util, cpu_util_irq, cpu_util_nice, cpu_util_sys, cpu_util_user%
dcgm.cpu_core.diagnostics.resultsCPU Core Diagnostics Results metrics.diag_cpu_eud_resultstate

Per exporter

These metrics refer to exporter/global instances.

Labels:

LabelDescription
jobjob label from exporter metrics.

Metrics:

MetricDescriptionDimensionsUnit
dcgm.exporter.health.statusExporter Health Status metrics.bind_unbind_eventstate
dcgm.exporter.inventory.softwareExporter Software and Firmware metrics.cuda_driver_version, driver_version, nvml_versionvalue

Troubleshooting

Debug Mode

Important: Debug mode is not supported for data collection jobs created via the UI using the Dyncfg feature.

To troubleshoot issues with the dcgm collector, run the go.d.plugin with the debug option enabled. The output should give you clues as to why the collector isn't working.

  • Navigate to the plugins.d directory, usually at /usr/libexec/netdata/plugins.d/. If that's not the case on your system, open netdata.conf and look for the plugins setting under [directories].

    cd /usr/libexec/netdata/plugins.d/
  • Switch to the netdata user.

    sudo -u netdata -s
  • Run the go.d.plugin to debug the collector:

    ./go.d.plugin -d -m dcgm

    To debug a specific job:

    ./go.d.plugin -d -m dcgm -j jobName

Getting Logs

If you're encountering problems with the dcgm collector, follow these steps to retrieve logs and identify potential issues:

  • Run the command specific to your system (systemd, non-systemd, or Docker container).
  • Examine the output for any warnings or error messages that might indicate issues. These messages should provide clues about the root cause of the problem.

System with systemd

Use the following command to view logs generated since the last Netdata service restart:

journalctl _SYSTEMD_INVOCATION_ID="$(systemctl show --value --property=InvocationID netdata)" --namespace=netdata --grep dcgm

System without systemd

Locate the collector log file, typically at /var/log/netdata/collector.log, and use grep to filter for collector's name:

grep dcgm /var/log/netdata/collector.log

Note: This method shows logs from all restarts. Focus on the latest entries for troubleshooting current issues.

Docker Container

If your Netdata runs in a Docker container named "netdata" (replace if different), use this command:

docker logs netdata 2>&1 | grep dcgm

Do you have any feedback for this page? If so, you can open a new issue on our netdata/learn repository.