expose health status as metric

components/event-publisher-proxy/cmd/event-publisher-proxy/main.go

@@ -58,6 +58,7 @@ func main() {
 	// metrics collector
 	metricsCollector := metrics.NewCollector(latency.NewBucketsProvider())
 	prometheus.MustRegister(metricsCollector)
+	metricsCollector.SetHealthStatus(true)
 
 	// Instantiate configured commander.
 	var c commander.Commander

components/event-publisher-proxy/pkg/metrics/collector.go

@@ -13,9 +13,14 @@ import (
 )
 
 const (
-	// BackendLatencyKey name of the backendLatencyHelp metric.
+	// HealthKey name of the health metric.
+	HealthKey = "eventing_epp_health"
+	// HealthHelp help text for the Health metric.
+	healthHelp = "The current health of the system. `1` indicates a healthy system"
+
+	// BackendLatencyKey name of the backendLatency metric.
 	BackendLatencyKey = "eventing_epp_backend_duration_milliseconds"
-	// backendLatencyHelp help text for the backendLatencyHelp metric.
+	// backendLatencyHelp help text for the backendLatency metric.
 	backendLatencyHelp = "The duration of sending events to the messaging server in milliseconds"
 
 	// durationKey name of the duration metric.
@@ -66,6 +71,8 @@ type Collector struct {
 	requests *prometheus.CounterVec
 
 	eventType *prometheus.CounterVec
+
+	health *prometheus.GaugeVec
 }
 
 // NewCollector creates a new instance of Collector.
@@ -102,7 +109,15 @@ func NewCollector(latency histogram.BucketsProvider) *Collector {
 				Name: RequestsKey,
 				Help: requestsHelp,
 			},
-			[]string{responseCodeLabel, methodLabel, pathLabel}),
+			[]string{responseCodeLabel, methodLabel, pathLabel},
+		),
+		health: prometheus.NewGaugeVec(
+			prometheus.GaugeOpts{
+				Name: HealthKey,
+				Help: healthHelp,
+			},
+			nil,
+		),
 	}
 }
 
@@ -112,6 +127,7 @@ func (c *Collector) Describe(ch chan<- *prometheus.Desc) {
 	c.eventType.Describe(ch)
 	c.requests.Describe(ch)
 	c.duration.Describe(ch)
+	c.health.Describe(ch)
 }
 
 // Collect implements the prometheus.Collector interface Collect method.
@@ -120,13 +136,23 @@ func (c *Collector) Collect(ch chan<- prometheus.Metric) {
 	c.eventType.Collect(ch)
 	c.requests.Collect(ch)
 	c.duration.Collect(ch)
+	c.health.Collect(ch)
 }
 
 // RecordLatency records a backendLatencyHelp metric.
 func (c *Collector) RecordBackendLatency(duration time.Duration, statusCode int, destSvc string) {
 	c.backendLatency.WithLabelValues(fmt.Sprint(statusCode), destSvc).Observe(float64(duration.Milliseconds()))
 }
 
+// SetHealthStatus updates the health metric.
+func (c *Collector) SetHealthStatus(healthy bool) {
+	var v float64
+	if healthy {
+		v = 1
+	}
+	c.health.WithLabelValues().Set(v)
+}
+
 // RecordEventType records an eventType metric.
 func (c *Collector) RecordEventType(eventType, eventSource string, statusCode int) {
 	c.eventType.WithLabelValues(eventType, eventSource, fmt.Sprint(statusCode)).Inc()

components/eventing-controller/Makefile

@@ -151,8 +151,10 @@ lint-thoroughly:
 
 update_metrics_docs:
 	@promlinter list -ojson ../event-publisher-proxy > metrics_epp.json
+	@jq 'sort_by(.Name)'  metrics_epp.json > metrics_epp_sorted.json
 	@promlinter list -ojson ../eventing-controller/ > metrics_ec.json
-	@gomplate -d ec=metrics_ec.json -d epp=metrics_epp.json -f hack/metrics.doc.tpl  | prettier --parser markdown > ../../docs/04-operation-guides/operations/evnt-02-eventing-metrics.md
+	@jq 'sort_by(.Name)'  metrics_ec.json > metrics_ec_sorted.json
+	@gomplate -d ec=metrics_ec_sorted.json -d epp=metrics_epp_sorted.json -f hack/metrics.doc.tpl  | prettier --parser markdown > ../../docs/04-operation-guides/operations/evnt-02-eventing-metrics.md
 
 update_docs: copy-crds
 	go run ../../hack/table-gen/main.go --crd-filename ../../installation/resources/crds/eventing/subscriptions.eventing.kyma-project.io.crd.yaml --md-filename ../../docs/05-technical-reference/00-custom-resources/evnt-01-subscription.md

docs/04-operation-guides/operations/evnt-02-eventing-metrics.md

@@ -9,18 +9,19 @@ The metrics follow the [Prometheus naming convention](https://prometheus.io/docs
 
 | Metric                                          | Description                                                                      |
 | ----------------------------------------------- | :------------------------------------------------------------------------------- |
-| **eventing_epp_requests_total**                 | The total number of requests                                                     |
 | **eventing_epp_backend_duration_milliseconds**  | The duration of sending events to the messaging server in milliseconds           |
 | **eventing_epp_event_type_published_total**     | The total number of events published for a given eventTypeLabel                  |
+| **eventing_epp_health**                         | The current health of the system. `1` indicates a healthy system                 |
 | **eventing_epp_requests_duration_milliseconds** | The duration of processing an incoming request (includes sending to the backend) |
+| **eventing_epp_requests_total**                 | The total number of requests                                                     |
 
 ### Metrics Emitted by Eventing Controller:
 
 | Metric                                                    | Description                                                                                                                 |
 | --------------------------------------------------------- | :-------------------------------------------------------------------------------------------------------------------------- |
+| **eventing_ec_event_type_subscribed_total**               | The total number of eventTypes subscribed using the Subscription CRD                                                        |
 | **eventing_ec_nats_delivery_per_subscription_total**      | The total number of dispatched events per subscription                                                                      |
 | **eventing_ec_nats_subscriber_dispatch_duration_seconds** | The duration of sending an incoming NATS message to the subscriber (not including processing the message in the dispatcher) |
-| **eventing_ec_event_type_subscribed_total**               | The total number of eventTypes subscribed using the Subscription CRD                                                        |
 
 ### Metrics Emitted by NATS Exporter:
 

resources/eventing/values.yaml

@@ -5,7 +5,7 @@ global:
   images:
     eventing_controller:
       name: eventing-controller
-      version: PR-17860
+      version: PR-17839
       directory: dev
       pullPolicy: "IfNotPresent"
     publisher_proxy: