Keystone Metric Improvements (#15706)

* execution duration is in seconds

* fixing histogram buckets for workflows

* adding triggerID to incrementRegisterTriggerFailureCounter

core/services/workflows/engine.go

@@ -28,7 +28,7 @@ import (
 )
 
 const (
-	fifteenMinutesMs             = 15 * 60 * 1000
+	fifteenMinutesSec            = 15 * 60
 	reservedFieldNameStepTimeout = "cre_step_timeout"
 	maxStepTimeoutOverrideSec    = 10 * 60 // 10 minutes
 )
@@ -446,7 +446,7 @@ func (e *Engine) registerTrigger(ctx context.Context, t *triggerCapability, trig
 	}
 	eventsCh, err := t.trigger.RegisterTrigger(ctx, triggerRegRequest)
 	if err != nil {
-		e.metrics.incrementRegisterTriggerFailureCounter(ctx)
+		e.metrics.with(platform.KeyTriggerID, triggerID).incrementRegisterTriggerFailureCounter(ctx)
 		// It's confusing that t.ID is different from triggerID, but
 		// t.ID is the capability ID, and triggerID is the trigger ID.
 		//
@@ -704,7 +704,7 @@ func (e *Engine) finishExecution(ctx context.Context, cma custmsg.MessageEmitter
 		e.metrics.updateWorkflowTimeoutDurationHistogram(ctx, executionDuration)
 	}
 
-	if executionDuration > fifteenMinutesMs {
+	if executionDuration > fifteenMinutesSec {
 		logCustMsg(ctx, cma, fmt.Sprintf("execution duration exceeded 15 minutes: %d (seconds)", executionDuration), l)
 		l.Warnf("execution duration exceeded 15 minutes: %d (seconds)", executionDuration)
 	}

core/services/workflows/monitoring.go

@@ -143,19 +143,19 @@ func MetricViews() []sdkmetric.View {
 		sdkmetric.NewView(
 			sdkmetric.Instrument{Name: "platform_engine_workflow_earlyexit_time_seconds"},
 			sdkmetric.Stream{Aggregation: sdkmetric.AggregationExplicitBucketHistogram{
-				Boundaries: []float64{0, 1, 10, 100},
+				Boundaries: []float64{0, 1, 10, 30, 120},
 			}},
 		),
 		sdkmetric.NewView(
 			sdkmetric.Instrument{Name: "platform_engine_workflow_completed_time_seconds"},
 			sdkmetric.Stream{Aggregation: sdkmetric.AggregationExplicitBucketHistogram{
-				Boundaries: []float64{0, 100, 1000, 10_000, 50_000, 100_0000, 500_000},
+				Boundaries: []float64{0, 10, 30, 60, 120, 300, 600, 900, 1200},
 			}},
 		),
 		sdkmetric.NewView(
 			sdkmetric.Instrument{Name: "platform_engine_workflow_error_time_seconds"},
 			sdkmetric.Stream{Aggregation: sdkmetric.AggregationExplicitBucketHistogram{
-				Boundaries: []float64{0, 20, 60, 120, 240},
+				Boundaries: []float64{0, 30, 60, 120, 240, 600},
 			}},
 		),
 		sdkmetric.NewView(