fix(monitor): register agent_nvidia_smi metrics

pkg/apis/compute/guests.go

@@ -1046,9 +1046,11 @@ type ServerHardwareInfoGPUPCIEInfo struct {
 type ServerHardwareInfoGPU struct {
 	*IsolatedDeviceModelHardwareInfo
 
+	// isolate device id
 	Id string `json:"id"`
 	// GPU model name
-	Model    string                  `json:"model"`
+	Model string `json:"model"`
+	// PCIE information
 	PCIEInfo *IsolatedDevicePCIEInfo `json:"pcie_info"`
 }
 

pkg/apis/compute/zz_generated.model.go

@@ -788,9 +788,9 @@ type SElasticcache struct {
 	NetworkId string `json:"network_id"`
 	// 带宽
 	Bandwidth int `json:"bandwidth"`
-	//  内网DNS
+	// 内网DNS
 	PrivateDNS string `json:"private_dns"`
-	//  内网IP地址
+	// 内网IP地址
 	PrivateIpAddr string `json:"private_ip_addr"`
 	// 内网访问端口
 	PrivateConnectPort int `json:"private_connect_port"`
@@ -1878,9 +1878,9 @@ type SLoadbalancerHealthCheck struct {
 	// 健康检查路径 /
 	HealthCheckHttpCode string `json:"health_check_http_code"`
 	HealthCheckRise     int    `json:"health_check_rise"`
-	//  健康检查健康阈值 3秒
+	// 健康检查健康阈值 3秒
 	HealthCheckFall int `json:"health_check_fall"`
-	//  健康检查不健康阈值 15秒
+	// 健康检查不健康阈值 15秒
 	HealthCheckTimeout int `json:"health_check_timeout"`
 	// 健康检查超时时间 10秒
 	HealthCheckInterval int    `json:"health_check_interval"`
@@ -2511,7 +2511,7 @@ type SSchedtag struct {
 	apis.SStandaloneResourceBase
 	apis.SScopedResourceBase
 	DefaultStrategy string `json:"default_strategy"`
-	// Column(VARCHAR(16, charset='ascii'), nullable=True, default='')
+	// Column(VARCHAR(16, charset='ascii'), nullable=True, default=”)
 	ResourceType string `json:"resource_type"`
 }
 

pkg/monitor/dbinit/metric_dbinit.go

@@ -389,6 +389,19 @@ func init() {
 			newMetricFieldCreateInput("temp_c", "Disk device temperature ", "", 1),
 		})
 
+	// agent nvidia_smi
+	RegistryMetricCreateInput("agent_nvidia_smi", "Collect Nvidia GPU metrics",
+		monitor.METRIC_RES_TYPE_AGENT, monitor.METRIC_DATABASE_TELE, 8, []monitor.MetricFieldCreateInput{
+			newMetricFieldCreateInput("clocks_current_graphics", "GPU current clocks, MHz", "", 1),
+			newMetricFieldCreateInput("clocks_current_memory", "GPU current memory clocks, MHz", "", 2),
+			newMetricFieldCreateInput("temperature_gpu", "GPU temperature", "", 3),
+			newMetricFieldCreateInput("memory_total", "GPU memory total size", "", 4),
+			newMetricFieldCreateInput("memory_free", "GPU memory free size", "", 5),
+			newMetricFieldCreateInput("memory_used", "GPU memory used size", "", 6),
+			newMetricFieldCreateInput("utilization_gpu", "GPU utilization", "", 7),
+			newMetricFieldCreateInput("utilization_memory", "GPU memory utilization", "", 8),
+		})
+
 	RegistryMetricCreateInput("storage", "Storage usage",
 		monitor.METRIC_RES_TYPE_STORAGE, monitor.METRIC_DATABASE_TELE, 1, []monitor.MetricFieldCreateInput{
 			newMetricFieldCreateInput("usage_active", "Storage utilization rate", monitor.METRIC_UNIT_PERCENT, 1),