Prometheus Receiver: Print a more informative message about 'up' metr…

…ic value (#1826) * Prometheus Receiver: Cleanup special handling of the 'up' metric * gofmt
open-telemetry · Sep 23, 2020 · 3c32f36 · 3c32f36
1 parent eb21271
commit 3c32f36
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 32 deletions.
diff --git a/receiver/prometheusreceiver/internal/metricsbuilder.go b/receiver/prometheusreceiver/internal/metricsbuilder.go
@@ -30,16 +30,11 @@ import (
 )
 
 const (
-	metricsSuffixCount      = "_count"
-	metricsSuffixBucket     = "_bucket"
-	metricsSuffixSum        = "_sum"
-	startTimeMetricName     = "process_start_time_seconds"
-	scrapeLatencyMetricName = "scrape_duration_seconds"
-	scrapeStatusMetricName  = "up"
-	scrapeStatusOk          = "200"
-	// The 'up' metric only reports whether or not the scrape succeeded - in the case that
-	// it fails, we set the status to '404', which is the most generic failure status.
-	scrapeStatusErr = "404"
+	metricsSuffixCount  = "_count"
+	metricsSuffixBucket = "_bucket"
+	metricsSuffixSum    = "_sum"
+	startTimeMetricName = "process_start_time_seconds"
+	scrapeUpMetricName  = "up"
 )
 
 var (
@@ -61,8 +56,6 @@ type metricBuilder struct {
 	useStartTimeMetric   bool
 	startTimeMetricRegex *regexp.Regexp
 	startTime            float64
-	scrapeLatencyMs      float64
-	scrapeStatus         string
 	logger               *zap.Logger
 	currentMf            MetricFamily
 }
@@ -106,16 +99,19 @@ func (b *metricBuilder) AddDataPoint(ls labels.Labels, t int64, v float64) error
 		b.hasInternalMetric = true
 		lm := ls.Map()
 		delete(lm, model.MetricNameLabel)
-		switch metricName {
-		case scrapeStatusMetricName:
-			if v == 1.0 {
-				b.scrapeStatus = scrapeStatusOk
+		// See https://www.prometheus.io/docs/concepts/jobs_instances/#automatically-generated-labels-and-time-series
+		// up: 1 if the instance is healthy, i.e. reachable, or 0 if the scrape failed.
+		if metricName == scrapeUpMetricName && v != 1.0 {
+			if v == 0.0 {
+				b.logger.Warn("Failed to scrape Prometheus endpoint",
+					zap.Int64("scrape_timestamp", t),
+					zap.String("target_labels", fmt.Sprintf("%v", lm)))
 			} else {
-				b.scrapeStatus = scrapeStatusErr
-				b.logger.Warn("http client error", zap.Int64("timestamp", t), zap.Float64("value", v), zap.String("labels", fmt.Sprintf("%v", lm)))
+				b.logger.Warn("The 'up' metric contains invalid value",
+					zap.Float64("value", v),
+					zap.Int64("scrape_timestamp", t),
+					zap.String("target_labels", fmt.Sprintf("%v", lm)))
 			}
-		case scrapeLatencyMetricName:
-			b.scrapeLatencyMs = v * 1000
 		}
 		return nil
 	case b.useStartTimeMetric && b.matchStartTimeMetric(metricName):
@@ -302,7 +298,7 @@ func timestampFromMs(timeAtMs int64) *timestamppb.Timestamp {
 }
 
 func isInternalMetric(metricName string) bool {
-	if metricName == "up" || strings.HasPrefix(metricName, "scrape_") {
+	if metricName == scrapeUpMetricName || strings.HasPrefix(metricName, "scrape_") {
 		return true
 	}
 	return false

diff --git a/receiver/prometheusreceiver/internal/transaction.go b/receiver/prometheusreceiver/internal/transaction.go
@@ -27,9 +27,6 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/prometheus/prometheus/pkg/labels"
 	"github.com/prometheus/prometheus/storage"
-	"go.opencensus.io/plugin/ochttp"
-	"go.opencensus.io/stats"
-	"go.opencensus.io/tag"
 	"go.uber.org/zap"
 	"google.golang.org/protobuf/types/known/timestamppb"
 
@@ -159,14 +156,6 @@ func (tr *transaction) Commit() error {
 		return err
 	}
 
-	if tr.metricBuilder.hasInternalMetric {
-		m := ochttp.ClientRoundtripLatency.M(tr.metricBuilder.scrapeLatencyMs)
-		stats.RecordWithTags(tr.ctx, []tag.Mutator{
-			tag.Upsert(ochttp.KeyClientStatus, tr.metricBuilder.scrapeStatus),
-		}, m)
-
-	}
-
 	if tr.useStartTimeMetric {
 		// AdjustStartTime - startTime has to be non-zero in this case.
 		if tr.metricBuilder.startTime == 0.0 {