PMM-10271 Fix meta-metric label for general collector (#515)

* PMM-10271 Fix meta-metric label for general collector

Meta-metric generated for general collector were incorrectly labeled as
dbstats. When dbstats and general collector were executed for a scrape request,
two metric were generated with the same name and labels. This duplicate metric
would trigger a panic in the metric generation library. This change updates the
label for general collector and fixes the issue.

* Fix tests for general collector

* Fix test for replication status collector

Author: Shashank Sinha

exporter/general_collector.go

@@ -49,7 +49,7 @@ func (d *generalCollector) Collect(ch chan<- prometheus.Metric) {
 }
 
 func (d *generalCollector) collect(ch chan<- prometheus.Metric) {
-	defer prometheus.MeasureCollectTime(ch, "mongodb", "dbstats")()
+	defer prometheus.MeasureCollectTime(ch, "mongodb", "general")()
 	ch <- mongodbUpMetric(d.ctx, d.base.client, d.base.logger)
 }
 

exporter/general_collector_test.go

@@ -37,15 +37,22 @@ func TestGeneralCollector(t *testing.T) {
 	client := tu.DefaultTestClient(ctx, t)
 	c := newGeneralCollector(ctx, client, logrus.New())
 
+	filter := []string{
+		"collector_scrape_time_ms",
+	}
+	count := testutil.CollectAndCount(c, filter...)
+	assert.Equal(t, len(filter), count, "Meta-metric for collector is missing")
+
 	// The last \n at the end of this string is important
 	expected := strings.NewReader(`
 	# HELP mongodb_up Whether MongoDB is up.
 	# TYPE mongodb_up gauge
 	mongodb_up 1
-	# HELP collector_scrape_time_ms Time taken for scrape by collector
-    # TYPE collector_scrape_time_ms gauge
-    collector_scrape_time_ms{collector="dbstats",exporter="mongodb"} 0` + "\n")
-	err := testutil.CollectAndCompare(c, expected)
+	` + "\n")
+	filter = []string{
+		"mongodb_up",
+	}
+	err := testutil.CollectAndCompare(c, expected, filter...)
 	require.NoError(t, err)
 
 	assert.NoError(t, client.Disconnect(ctx))
@@ -54,9 +61,10 @@ func TestGeneralCollector(t *testing.T) {
 	# HELP mongodb_up Whether MongoDB is up.
 	# TYPE mongodb_up gauge
 	mongodb_up 0
-	# HELP collector_scrape_time_ms Time taken for scrape by collector
-    # TYPE collector_scrape_time_ms gauge
-    collector_scrape_time_ms{collector="dbstats",exporter="mongodb"} 0` + "\n")
-	err = testutil.CollectAndCompare(c, expected)
+	` + "\n")
+	filter = []string{
+		"mongodb_up",
+	}
+	err = testutil.CollectAndCompare(c, expected, filter...)
 	require.NoError(t, err)
 }

exporter/indexstats_collector_test.go

@@ -79,7 +79,6 @@ mongodb_indexstats_accesses_ops{collection="testcol_02",database="testdb",key_na
 mongodb_indexstats_accesses_ops{collection="testcol_02",database="testdb",key_name="idx_01"} 0` +
 		"\n")
 
-	// on race run we will have collector_scrape_time_ms > 0, in normal run it will be fast and will 0
 	filter := []string{
 		"mongodb_indexstats_accesses_ops",
 	}
@@ -132,7 +131,6 @@ func TestDescendingIndexOverride(t *testing.T) {
   mongodb_indexstats_accesses_ops{collection="testcol_02",database="testdb",key_name="f1_1"} 0
   mongodb_indexstats_accesses_ops{collection="testcol_02",database="testdb",key_name="f1_DESC"} 0` + "\n")
 
-	// on race run we will have collector_scrape_time_ms > 0, in normal run it will be fast and will 0
 	filter := []string{
 		"mongodb_indexstats_accesses_ops",
 	}

exporter/replset_status_collector_test.go

@@ -69,10 +69,9 @@ func TestReplsetStatusCollectorNoSharding(t *testing.T) {
 
 	c := newReplicationSetStatusCollector(ctx, client, logrus.New(), false, ti)
 
-	expected := strings.NewReader(`
-	# HELP collector_scrape_time_ms Time taken for scrape by collector
-    # TYPE collector_scrape_time_ms gauge
-    collector_scrape_time_ms{collector="replset_status",exporter="mongodb"} 0` + "\n")
-	err := testutil.CollectAndCompare(c, expected)
-	assert.NoError(t, err)
+	// Replication set metrics should not be generated for unsharded server
+	count := testutil.CollectAndCount(c)
+
+	metaMetricCount := 1
+	assert.Equal(t, metaMetricCount, count, "Mismatch in metric count for collector run on unsharded server")
 }