Merge pull request #58192 from ravisantoshgudimetla/premeptions-metrics-additions

pkg/scheduler/metrics/metrics.go

@@ -59,6 +59,14 @@ var (
 			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
 		},
 	)
+	SchedulingAlgorithmPremptionEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: schedulerSubsystem,
+			Name:      "scheduling_algorithm_preemption_evaluation",
+			Help:      "Scheduling algorithm preemption evaluation duration",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
 	BindingLatency = prometheus.NewHistogram(
 		prometheus.HistogramOpts{
 			Subsystem: schedulerSubsystem,
@@ -67,6 +75,18 @@ var (
 			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
 		},
 	)
+	PreemptionVictims = prometheus.NewGauge(
+		prometheus.GaugeOpts{
+			Subsystem: schedulerSubsystem,
+			Name:      "pod_preemption_victims",
+			Help:      "Number of selected preemption victims",
+		})
+	PreemptionAttempts = prometheus.NewCounter(
+		prometheus.CounterOpts{
+			Subsystem: schedulerSubsystem,
+			Name:      "total_preemption_attempts",
+			Help:      "Total preemption attempts in the cluster till now",
+		})
 )
 
 var registerMetrics sync.Once
@@ -78,8 +98,12 @@ func Register() {
 		prometheus.MustRegister(E2eSchedulingLatency)
 		prometheus.MustRegister(SchedulingAlgorithmLatency)
 		prometheus.MustRegister(BindingLatency)
+
 		prometheus.MustRegister(SchedulingAlgorithmPredicateEvaluationDuration)
 		prometheus.MustRegister(SchedulingAlgorithmPriorityEvaluationDuration)
+		prometheus.MustRegister(SchedulingAlgorithmPremptionEvaluationDuration)
+		prometheus.MustRegister(PreemptionVictims)
+		prometheus.MustRegister(PreemptionAttempts)
 	})
 }
 

pkg/scheduler/scheduler.go

@@ -216,7 +216,9 @@ func (sched *Scheduler) preempt(preemptor *v1.Pod, scheduleErr error) (string, e
 		glog.Errorf("Error getting the updated preemptor pod object: %v", err)
 		return "", err
 	}
+
 	node, victims, nominatedPodsToClear, err := sched.config.Algorithm.Preempt(preemptor, sched.config.NodeLister, scheduleErr)
+	metrics.PreemptionVictims.Set(float64(len(victims)))
 	if err != nil {
 		glog.Errorf("Error preempting victims to make room for %v/%v.", preemptor.Namespace, preemptor.Name)
 		return "", err
@@ -440,18 +442,20 @@ func (sched *Scheduler) scheduleOne() {
 	// Synchronously attempt to find a fit for the pod.
 	start := time.Now()
 	suggestedHost, err := sched.schedule(pod)
-	metrics.SchedulingAlgorithmLatency.Observe(metrics.SinceInMicroseconds(start))
 	if err != nil {
 		// schedule() may have failed because the pod would not fit on any host, so we try to
 		// preempt, with the expectation that the next time the pod is tried for scheduling it
 		// will fit due to the preemption. It is also possible that a different pod will schedule
 		// into the resources that were preempted, but this is harmless.
 		if fitError, ok := err.(*core.FitError); ok {
+			preemptionStartTime := time.Now()
 			sched.preempt(pod, fitError)
+			metrics.PreemptionAttempts.Inc()
+			metrics.SchedulingAlgorithmPremptionEvaluationDuration.Observe(metrics.SinceInMicroseconds(preemptionStartTime))
 		}
 		return
 	}
-
+	metrics.SchedulingAlgorithmLatency.Observe(metrics.SinceInMicroseconds(start))
 	// Tell the cache to assume that a pod now is running on a given node, even though it hasn't been bound yet.
 	// This allows us to keep scheduling without waiting on binding to occur.
 	assumedPod := pod.DeepCopy()