feat: add audit_last_run_end_time metric

pkg/audit/manager.go

@@ -161,10 +161,14 @@ func (am *Manager) audit(ctx context.Context) error {
 	// record audit latency
 	defer func() {
 		logFinish(am.log)
-		latency := time.Since(startTime)
+		endTime := time.Now()
+		latency := endTime.Sub(startTime)
 		if err := am.reporter.reportLatency(latency); err != nil {
 			am.log.Error(err, "failed to report latency")
 		}
+		if err := am.reporter.reportRunEnd(endTime); err != nil {
+			am.log.Error(err, "failed to report run end time")
+		}
 	}()
 
 	if err := am.reporter.reportRunStart(startTime); err != nil {

pkg/audit/stats_reporter.go

@@ -12,15 +12,17 @@ import (
 )
 
 const (
-	violationsMetricName    = "violations"
-	auditDurationMetricName = "audit_duration_seconds"
-	lastRunTimeMetricName   = "audit_last_run_time"
+	violationsMetricName       = "violations"
+	auditDurationMetricName    = "audit_duration_seconds"
+	lastRunStartTimeMetricName = "audit_last_run_time"
+	lastRunEndTimeMetricName   = "audit_last_run_end_time"
 )
 
 var (
-	violationsM    = stats.Int64(violationsMetricName, "Total number of audited violations", stats.UnitDimensionless)
-	auditDurationM = stats.Float64(auditDurationMetricName, "Latency of audit operation in seconds", stats.UnitSeconds)
-	lastRunTimeM   = stats.Float64(lastRunTimeMetricName, "Timestamp of last audit run time", stats.UnitSeconds)
+	violationsM       = stats.Int64(violationsMetricName, "Total number of audited violations", stats.UnitDimensionless)
+	auditDurationM    = stats.Float64(auditDurationMetricName, "Latency of audit operation in seconds", stats.UnitSeconds)
+	lastRunStartTimeM = stats.Float64(lastRunStartTimeMetricName, "Timestamp of last audit run starting time", stats.UnitSeconds)
+	lastRunEndTimeM   = stats.Float64(lastRunEndTimeMetricName, "Timestamp of last audit run ending time", stats.UnitSeconds)
 
 	enforcementActionKey = tag.MustNewKey("enforcement_action")
 )
@@ -45,9 +47,13 @@ func register() error {
 			Aggregation: view.Distribution(1*60, 3*60, 5*60, 10*60, 15*60, 20*60, 40*60, 80*60, 160*60, 320*60),
 		},
 		{
-			Name:        lastRunTimeMetricName,
-			Measure:     lastRunTimeM,
-			Description: "Timestamp of last audit run time",
+			Name:        lastRunStartTimeMetricName,
+			Measure:     lastRunStartTimeM,
+			Aggregation: view.LastValue(),
+		},
+		{
+			Name:        lastRunEndTimeMetricName,
+			Measure:     lastRunEndTimeM,
 			Aggregation: view.LastValue(),
 		},
 	}
@@ -80,8 +86,18 @@ func (r *reporter) reportRunStart(t time.Time) error {
 		return err
 	}
 
-	val := float64(t.UnixNano()) / 1e9
-	return metrics.Record(ctx, lastRunTimeM.M(val))
+	val := float64(t.Unix())
+	return metrics.Record(ctx, lastRunStartTimeM.M(val))
+}
+
+func (r *reporter) reportRunEnd(t time.Time) error {
+	ctx, err := tag.New(context.Background())
+	if err != nil {
+		return err
+	}
+
+	val := float64(t.Unix())
+	return metrics.Record(ctx, lastRunEndTimeM.M(val))
 }
 
 // newStatsReporter creates a reporter for audit metrics.

pkg/audit/stats_reporter_test.go

@@ -101,30 +101,50 @@ func checkData(t *testing.T, name string, wantRowLength int) *view.Row {
 }
 
 func TestLastRestartCheck(t *testing.T) {
-	wantTime := time.Now()
-	wantTs := float64(wantTime.UnixNano()) / 1e9
+	wantStartTime := time.Now()
+	wantEndTime := wantStartTime.Add(1 * time.Minute)
+	wantStartTs := float64(wantStartTime.Unix())
+	wantEndTs := float64(wantEndTime.Unix())
 	const wantRowLength = 1
 
 	r, err := newStatsReporter()
 	if err != nil {
 		t.Fatalf("got newStatsReporter() error %v", err)
 	}
 
-	err = r.reportRunStart(wantTime)
+	err = r.reportRunStart(wantStartTime)
 	if err != nil {
 		t.Fatalf("reportRunStart error %v", err)
 	}
-	row := checkData(t, lastRunTimeMetricName, wantRowLength)
+	row := checkData(t, lastRunStartTimeMetricName, wantRowLength)
 	got, ok := row.Data.(*view.LastValueData)
 	if !ok {
-		t.Error("lastRunTimeMetricName should have aggregation LastValue()")
+		t.Errorf("%s should have aggregation LastValue()", lastRunStartTimeMetricName)
 	}
 
 	if len(row.Tags) != 0 {
-		t.Errorf("got %q tags %v, want empty", lastRunTimeMetricName, row.Tags)
+		t.Errorf("got %q tags %v, want empty", lastRunStartTimeMetricName, row.Tags)
 	}
 
-	if got.Value != wantTs {
-		t.Errorf("got %q = %v, want %v", lastRunTimeMetricName, got.Value, wantTs)
+	if got.Value != wantStartTs {
+		t.Errorf("got %q = %v, want %v", lastRunStartTimeMetricName, got.Value, wantStartTs)
+	}
+
+	err = r.reportRunEnd(wantEndTime)
+	if err != nil {
+		t.Fatalf("reportRunEnd error %v", err)
+	}
+	row = checkData(t, lastRunEndTimeMetricName, wantRowLength)
+	got, ok = row.Data.(*view.LastValueData)
+	if !ok {
+		t.Errorf("%s should have aggregation LastValue()", lastRunEndTimeMetricName)
+	}
+
+	if len(row.Tags) != 0 {
+		t.Errorf("got %q tags %v, want empty", lastRunEndTimeMetricName, row.Tags)
+	}
+
+	if got.Value != wantEndTs {
+		t.Errorf("got %q = %v, want %v", lastRunEndTimeMetricName, got.Value, wantEndTs)
 	}
 }

website/docs/audit.md

@@ -13,7 +13,8 @@ There are three ways to gather audit results, depending on the level of detail n
 
 Prometheus metrics provide an aggregated look at the number of audit violations:
 
-* `gatekeeper_audit_last_run_time` provides the timestamp of the most recently completed audit run
+* `gatekeeper_audit_last_run_time` provides the start time timestamp of the most recent audit run
+* `gatekeeper_audit_last_run_end_time` provides the end time timestamp of the last completed audit run
 * `gatekeeper_violations` provides the total number of audited violations for the last audit run, broken down by violation severity
 
 ### Constraint Status

website/docs/metrics.md

@@ -113,7 +113,13 @@ Below are the list of metrics provided by Gatekeeper:
 
 - Name: `gatekeeper_audit_last_run_time`
 
-    Description: `Timestamp of last audit run time`
+    Description: `Timestamp of last audit run starting time`
+
+    Aggregation: `LastValue`
+
+- Name: `gatekeeper_audit_last_run_end_time`
+
+    Description: `Timestamp of last audit run ending time`
 
     Aggregation: `LastValue`
 

website/versioned_docs/version-v3.6.x/audit.md

@@ -13,7 +13,7 @@ There are three ways to gather audit results, depending on the level of detail n
 
 Prometheus metrics provide an aggregated look at the number of audit violations:
 
-* `gatekeeper_audit_last_run_time` provides the timestamp of the most recently completed audit run
+* `gatekeeper_audit_last_run_time` provides the start time timestamp of the most recent audit run
 * `gatekeeper_violations` provides the total number of audited violations for the last audit run, broken down by violation severity
 
 ### Constraint Status

website/versioned_docs/version-v3.7.x/audit.md

@@ -13,7 +13,7 @@ There are three ways to gather audit results, depending on the level of detail n
 
 Prometheus metrics provide an aggregated look at the number of audit violations:
 
-* `gatekeeper_audit_last_run_time` provides the timestamp of the most recently completed audit run
+* `gatekeeper_audit_last_run_time` provides the start time timestamp of the most recent audit run
 * `gatekeeper_violations` provides the total number of audited violations for the last audit run, broken down by violation severity
 
 ### Constraint Status

website/versioned_docs/version-v3.8.x/audit.md

@@ -13,7 +13,7 @@ There are three ways to gather audit results, depending on the level of detail n
 
 Prometheus metrics provide an aggregated look at the number of audit violations:
 
-* `gatekeeper_audit_last_run_time` provides the timestamp of the most recently completed audit run
+* `gatekeeper_audit_last_run_time` provides the start time timestamp of the most recent audit run
 * `gatekeeper_violations` provides the total number of audited violations for the last audit run, broken down by violation severity
 
 ### Constraint Status

website/versioned_docs/version-v3.9.x/audit.md

@@ -13,7 +13,7 @@ There are three ways to gather audit results, depending on the level of detail n
 
 Prometheus metrics provide an aggregated look at the number of audit violations:
 
-* `gatekeeper_audit_last_run_time` provides the timestamp of the most recently completed audit run
+* `gatekeeper_audit_last_run_time` provides the start time timestamp of the most recent audit run
 * `gatekeeper_violations` provides the total number of audited violations for the last audit run, broken down by violation severity
 
 ### Constraint Status