feat: collect model performance metrics (#4635)

controller/perf_metrics.go

@@ -0,0 +1,46 @@
+package controller
+
+import (
+	"net/http"
+	"strconv"
+
+	perfmetrics "github.com/QuantumNous/new-api/pkg/perf_metrics"
+
+	"github.com/gin-gonic/gin"
+)
+
+func GetPerfMetrics(c *gin.Context) {
+	modelName := c.Query("model")
+	if modelName == "" {
+		c.JSON(http.StatusBadRequest, gin.H{
+			"success": false,
+			"message": "model is required",
+		})
+		return
+	}
+
+	hours := 24
+	if rawHours := c.Query("hours"); rawHours != "" {
+		if parsed, err := strconv.Atoi(rawHours); err == nil {
+			hours = parsed
+		}
+	}
+
+	result, err := perfmetrics.Query(perfmetrics.QueryParams{
+		Model: modelName,
+		Group: c.Query("group"),
+		Hours: hours,
+	})
+	if err != nil {
+		c.JSON(http.StatusInternalServerError, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"data":    result,
+	})
+}

controller/relay.go

@@ -15,6 +15,7 @@ import (
 	"github.com/QuantumNous/new-api/logger"
 	"github.com/QuantumNous/new-api/middleware"
 	"github.com/QuantumNous/new-api/model"
+	perfmetrics "github.com/QuantumNous/new-api/pkg/perf_metrics"
 	"github.com/QuantumNous/new-api/relay"
 	relaycommon "github.com/QuantumNous/new-api/relay/common"
 	relayconstant "github.com/QuantumNous/new-api/relay/constant"
@@ -239,6 +240,11 @@ func Relay(c *gin.Context, relayFormat types.RelayFormat) {
 		retryLogStr := fmt.Sprintf("重试：%s", strings.Trim(strings.Join(strings.Fields(fmt.Sprint(useChannel)), "->"), "[]"))
 		logger.LogInfo(c, retryLogStr)
 	}
+	if newAPIError != nil {
+		gopool.Go(func() {
+			perfmetrics.RecordRelaySample(relayInfo, false)
+		})
+	}
 }
 
 var upgrader = websocket.Upgrader{