fix: metric check sp

bnb-chain · Jun 3, 2024 · 834e1f4 · 834e1f4
1 parent ac790ef
commit 834e1f4
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 4 deletions.
diff --git a/base/gfspvgmgr/virtual_group_manager.go b/base/gfspvgmgr/virtual_group_manager.go
@@ -806,7 +806,7 @@ func (checker *HealthChecker) checkSPHealth(sp *sptypes.StorageProvider) bool {
 
 		resp, err := client.Do(req)
 		duration := time.Since(start)
-		metrics.SPHealthCheckerDuration.WithLabelValues(SPHealthCheckerDuration, util.Uint32ToString(sp.Id)).Observe(duration.Seconds())
+		metrics.SPHealthCheckerTime.WithLabelValues(SPHealthCheckerDuration, util.Uint32ToString(sp.Id)).Observe(duration.Seconds())
 		if err != nil {
 			log.CtxErrorw(context.Background(), "failed to connect to sp", "sp", sp, "error", err, "duration", duration)
 			time.Sleep(defaultSPHealthCheckerRetryInterval)
@@ -818,7 +818,7 @@ func (checker *HealthChecker) checkSPHealth(sp *sptypes.StorageProvider) bool {
 			log.CtxInfow(context.Background(), "succeed to check the sp healthy", "sp", sp, "duration", duration)
 			return true
 		} else {
-			metrics.SPHealthCheckerFailure.WithLabelValues(SPHealthCheckerFailure, util.Uint32ToString(sp.Id)).Inc()
+			metrics.SPHealthCheckerFailureCounter.WithLabelValues(SPHealthCheckerFailure, util.Uint32ToString(sp.Id)).Inc()
 			log.CtxErrorw(context.Background(), "failed to check sp healthy", "sp", sp, "http_status_code", resp.StatusCode, "duration", duration)
 			time.Sleep(defaultSPHealthCheckerRetryInterval)
 		}

diff --git a/pkg/metrics/metric_items.go b/pkg/metrics/metric_items.go
@@ -61,6 +61,8 @@ var MetricsItems = []prometheus.Collector{
 	ManagerCounter,
 	ManagerTime,
 	GCBlockNumberGauge,
+	SPHealthCheckerTime,
+	SPHealthCheckerFailureCounter,
 
 	// workflow metrics category
 	PerfApprovalTime,
@@ -247,14 +249,14 @@ var (
 		Name: "gc_block_number",
 		Help: "Track the next gc block number.",
 	}, []string{"gc_block_number"})
-	SPHealthCheckerDuration = prometheus.NewHistogramVec(
+	SPHealthCheckerTime = prometheus.NewHistogramVec(
 		prometheus.HistogramOpts{
 			Name: "request_time",
 			Help: "Request duration in seconds.",
 		},
 		[]string{"request_time", "storage_provider"},
 	)
-	SPHealthCheckerFailure = prometheus.NewCounterVec(
+	SPHealthCheckerFailureCounter = prometheus.NewCounterVec(
 		prometheus.CounterOpts{
 			Name: "request_qps",
 			Help: "Request failure count.",