fix: recover event when prom_for_duration is 0

alert/process/process.go

@@ -390,6 +390,13 @@ func (p *Processor) handleEvent(events []*models.AlertCurEvent) {
 		if event == nil {
 			continue
 		}
+
+		if _, has := p.pendingsUseByRecover.Get(event.Hash); has {
+			p.pendingsUseByRecover.UpdateLastEvalTime(event.Hash, event.LastEvalTime)
+		} else {
+			p.pendingsUseByRecover.Set(event.Hash, event)
+		}
+
 		if p.rule.PromForDuration == 0 {
 			fireEvents = append(fireEvents, event)
 			if severity > event.Severity {
@@ -398,15 +405,13 @@ func (p *Processor) handleEvent(events []*models.AlertCurEvent) {
 			continue
 		}
 
-		var preTriggerTime int64
+		var preTriggerTime int64 // 第一个 pending event 的触发时间
 		preEvent, has := p.pendings.Get(event.Hash)
 		if has {
 			p.pendings.UpdateLastEvalTime(event.Hash, event.LastEvalTime)
-			p.pendingsUseByRecover.UpdateLastEvalTime(event.Hash, event.LastEvalTime)
 			preTriggerTime = preEvent.TriggerTime
 		} else {
 			p.pendings.Set(event.Hash, event)
-			p.pendingsUseByRecover.Set(event.Hash, event)
 			preTriggerTime = event.TriggerTime
 		}
 

models/prom_alert_rule_test.go

@@ -21,7 +21,7 @@ func TestConvertAlert(t *testing.T) {
 		t.Errorf("Failed to Unmarshal, err: %s", err)
 	}
 	t.Logf("jobMissing: %+v", jobMissing[0])
-	convJobMissing := models.ConvertAlert(jobMissing[0], "30s")
+	convJobMissing := models.ConvertAlert(jobMissing[0], "30s", "1", 0)
 	if convJobMissing.PromEvalInterval != 30 {
 		t.Errorf("PromEvalInterval is expected to be 30, but got %d",
 			convJobMissing.PromEvalInterval)
@@ -45,7 +45,7 @@ func TestConvertAlert(t *testing.T) {
       description: "Prometheus rule evaluation took more time than the scheduled interval. It indicates a slower storage backend access or too complex query.\n  VALUE = {{ $value }}\n  LABELS = {{ $labels }}"
 `), &ruleEvaluationSlow)
 	t.Logf("ruleEvaluationSlow: %+v", ruleEvaluationSlow[0])
-	convRuleEvaluationSlow := models.ConvertAlert(ruleEvaluationSlow[0], "1m")
+	convRuleEvaluationSlow := models.ConvertAlert(ruleEvaluationSlow[0], "1m", "1", 0)
 	if convRuleEvaluationSlow.PromEvalInterval != 60 {
 		t.Errorf("PromEvalInterval is expected to be 60, but got %d",
 			convJobMissing.PromEvalInterval)
@@ -69,7 +69,7 @@ func TestConvertAlert(t *testing.T) {
       description: "A Prometheus target has disappeared. An exporter might be crashed.\n  VALUE = {{ $value }}\n  LABELS = {{ $labels }}"
 `), &targetMissing)
 	t.Logf("targetMissing: %+v", targetMissing[0])
-	convTargetMissing := models.ConvertAlert(targetMissing[0], "1h")
+	convTargetMissing := models.ConvertAlert(targetMissing[0], "1h", "1", 0)
 	if convTargetMissing.PromEvalInterval != 3600 {
 		t.Errorf("PromEvalInterval is expected to be 3600, but got %d",
 			convTargetMissing.PromEvalInterval)