sat: collect NVLink status/errors, fix nv-hostengine restart via systemd

Add NVLink port status (nvidia-smi nvlink -s), error counters
(nvidia-smi nvlink -e), and dcgmi nvlink status to the support bundle,
and enrich HardwarePCIeDevice entries with per-link telemetry.

Replace the raw nv-hostengine pkill/restart dance in bee-nvidia-load
with systemctl restart/start of nvidia-dcgm.service, and order
bee-nvidia.service Before= nvidia-dcgm.service and
nvidia-fabricmanager.service so modules/device nodes exist before
those units start.
This commit is contained in:
Mikhail Chusavitin
2026-07-07 10:21:56 +03:00
parent a98721b038
commit 33d6eee9cf
6 changed files with 339 additions and 81 deletions
+64 -49
View File
@@ -59,13 +59,13 @@ type RuntimeInterface struct {
}
type HardwareSnapshot struct {
Board HardwareBoard `json:"board"`
Firmware []HardwareFirmwareRecord `json:"firmware,omitempty"`
CPUs []HardwareCPU `json:"cpus,omitempty"`
Memory []HardwareMemory `json:"memory,omitempty"`
Storage []HardwareStorage `json:"storage,omitempty"`
PCIeDevices []HardwarePCIeDevice `json:"pcie_devices,omitempty"`
PowerSupplies []HardwarePowerSupply `json:"power_supplies,omitempty"`
Board HardwareBoard `json:"board"`
Firmware []HardwareFirmwareRecord `json:"firmware,omitempty"`
CPUs []HardwareCPU `json:"cpus,omitempty"`
Memory []HardwareMemory `json:"memory,omitempty"`
Storage []HardwareStorage `json:"storage,omitempty"`
PCIeDevices []HardwarePCIeDevice `json:"pcie_devices,omitempty"`
PowerSupplies []HardwarePowerSupply `json:"power_supplies,omitempty"`
Sensors *HardwareSensors `json:"sensors,omitempty"`
EventLogs []HardwareEventLog `json:"event_logs,omitempty"`
PlatformConfig *json.RawMessage `json:"platform_config,omitempty"`
@@ -178,48 +178,63 @@ type HardwareStorage struct {
type HardwarePCIeDevice struct {
HardwareComponentStatus
Slot *string `json:"slot,omitempty"`
VendorID *int `json:"vendor_id,omitempty"`
DeviceID *int `json:"device_id,omitempty"`
NUMANode *int `json:"numa_node,omitempty"`
TemperatureC *float64 `json:"temperature_c,omitempty"`
PowerW *float64 `json:"power_w,omitempty"`
LifeRemainingPct *float64 `json:"life_remaining_pct,omitempty"`
LifeUsedPct *float64 `json:"life_used_pct,omitempty"`
ECCCorrectedTotal *int64 `json:"ecc_corrected_total,omitempty"`
ECCUncorrectedTotal *int64 `json:"ecc_uncorrected_total,omitempty"`
HWSlowdown *bool `json:"hw_slowdown,omitempty"`
BatteryChargePct *float64 `json:"battery_charge_pct,omitempty"`
BatteryHealthPct *float64 `json:"battery_health_pct,omitempty"`
BatteryTemperatureC *float64 `json:"battery_temperature_c,omitempty"`
BatteryVoltageV *float64 `json:"battery_voltage_v,omitempty"`
BatteryReplaceRequired *bool `json:"battery_replace_required,omitempty"`
SFPPresent *bool `json:"sfp_present,omitempty"`
SFPIdentifier *string `json:"sfp_identifier,omitempty"`
SFPConnector *string `json:"sfp_connector,omitempty"`
SFPVendor *string `json:"sfp_vendor,omitempty"`
SFPPartNumber *string `json:"sfp_part_number,omitempty"`
SFPSerialNumber *string `json:"sfp_serial_number,omitempty"`
SFPWavelengthNM *float64 `json:"sfp_wavelength_nm,omitempty"`
SFPTemperatureC *float64 `json:"sfp_temperature_c,omitempty"`
SFPTXPowerDBM *float64 `json:"sfp_tx_power_dbm,omitempty"`
SFPRXPowerDBM *float64 `json:"sfp_rx_power_dbm,omitempty"`
SFPVoltageV *float64 `json:"sfp_voltage_v,omitempty"`
SFPBiasMA *float64 `json:"sfp_bias_ma,omitempty"`
BDF *string `json:"-"`
DeviceClass *string `json:"device_class,omitempty"`
Manufacturer *string `json:"manufacturer,omitempty"`
Model *string `json:"model,omitempty"`
LinkWidth *int `json:"link_width,omitempty"`
LinkSpeed *string `json:"link_speed,omitempty"`
MaxLinkWidth *int `json:"max_link_width,omitempty"`
MaxLinkSpeed *string `json:"max_link_speed,omitempty"`
SerialNumber *string `json:"serial_number,omitempty"`
Firmware *string `json:"firmware,omitempty"`
MacAddresses []string `json:"mac_addresses,omitempty"`
Present *bool `json:"present,omitempty"`
IOMMUGroup *int `json:"iommu_group,omitempty"`
Telemetry map[string]any `json:"-"`
Slot *string `json:"slot,omitempty"`
VendorID *int `json:"vendor_id,omitempty"`
DeviceID *int `json:"device_id,omitempty"`
NUMANode *int `json:"numa_node,omitempty"`
TemperatureC *float64 `json:"temperature_c,omitempty"`
PowerW *float64 `json:"power_w,omitempty"`
LifeRemainingPct *float64 `json:"life_remaining_pct,omitempty"`
LifeUsedPct *float64 `json:"life_used_pct,omitempty"`
ECCCorrectedTotal *int64 `json:"ecc_corrected_total,omitempty"`
ECCUncorrectedTotal *int64 `json:"ecc_uncorrected_total,omitempty"`
HWSlowdown *bool `json:"hw_slowdown,omitempty"`
BatteryChargePct *float64 `json:"battery_charge_pct,omitempty"`
BatteryHealthPct *float64 `json:"battery_health_pct,omitempty"`
BatteryTemperatureC *float64 `json:"battery_temperature_c,omitempty"`
BatteryVoltageV *float64 `json:"battery_voltage_v,omitempty"`
BatteryReplaceRequired *bool `json:"battery_replace_required,omitempty"`
SFPPresent *bool `json:"sfp_present,omitempty"`
SFPIdentifier *string `json:"sfp_identifier,omitempty"`
SFPConnector *string `json:"sfp_connector,omitempty"`
SFPVendor *string `json:"sfp_vendor,omitempty"`
SFPPartNumber *string `json:"sfp_part_number,omitempty"`
SFPSerialNumber *string `json:"sfp_serial_number,omitempty"`
SFPWavelengthNM *float64 `json:"sfp_wavelength_nm,omitempty"`
SFPTemperatureC *float64 `json:"sfp_temperature_c,omitempty"`
SFPTXPowerDBM *float64 `json:"sfp_tx_power_dbm,omitempty"`
SFPRXPowerDBM *float64 `json:"sfp_rx_power_dbm,omitempty"`
SFPVoltageV *float64 `json:"sfp_voltage_v,omitempty"`
SFPBiasMA *float64 `json:"sfp_bias_ma,omitempty"`
BDF *string `json:"-"`
DeviceClass *string `json:"device_class,omitempty"`
Manufacturer *string `json:"manufacturer,omitempty"`
Model *string `json:"model,omitempty"`
LinkWidth *int `json:"link_width,omitempty"`
LinkSpeed *string `json:"link_speed,omitempty"`
MaxLinkWidth *int `json:"max_link_width,omitempty"`
MaxLinkSpeed *string `json:"max_link_speed,omitempty"`
SerialNumber *string `json:"serial_number,omitempty"`
Firmware *string `json:"firmware,omitempty"`
MacAddresses []string `json:"mac_addresses,omitempty"`
Present *bool `json:"present,omitempty"`
IOMMUGroup *int `json:"iommu_group,omitempty"`
NVLinks []HardwareNVLinkPort `json:"nvlinks,omitempty"`
Telemetry map[string]any `json:"-"`
}
// HardwareNVLinkPort describes a single NVLink lane on a GPU, as reported by
// "nvidia-smi nvlink -s" (speed/active state) and "nvidia-smi nvlink -e"
// (per-link error counters). Only populated on GPU-class HardwarePCIeDevice
// entries. An inactive link is not necessarily a fault: some GPU SKUs/boards
// reserve lanes as standby failover paths by design.
type HardwareNVLinkPort struct {
Index int `json:"index"`
Active bool `json:"active"`
SpeedGBs *float64 `json:"speed_gbs,omitempty"`
ReplayErrors *int64 `json:"replay_errors,omitempty"`
RecoveryErrors *int64 `json:"recovery_errors,omitempty"`
CRCErrors *int64 `json:"crc_errors,omitempty"`
}
type HardwarePowerSupply struct {