JMX Scraper: reverted some metric name changes

JMX Metrics: metric descriptions updated to match common convention

Author: robsunday

jmx-metrics/src/integrationTest/java/io/opentelemetry/contrib/jmxmetrics/target_systems/KafkaIntegrationTest.java

@@ -165,25 +165,25 @@ List<Consumer<Metric>> kafkaBrokerAssertions() {
             assertGauge(
                 metric,
                 "kafka.controller.active.count",
-                "controller is active on broker",
+                "For KRaft mode, the number of active controllers in the cluster. For ZooKeeper, indicates whether the broker is the controller broker.",
                 "{controller}"),
         metric ->
             assertSum(
                 metric,
                 "kafka.leader.election.rate",
-                "leader election rate - increasing indicates broker failures",
+                "Leader election rate - increasing indicates broker failures",
                 "{election}"),
         metric ->
             assertGauge(
                 metric,
                 "kafka.max.lag",
-                "max lag in messages between follower and leader replicas",
+                "Max lag in messages between follower and leader replicas",
                 "{message}"),
         metric ->
             assertSum(
                 metric,
                 "kafka.unclean.election.rate",
-                "unclean leader election rate - increasing indicates broker failures",
+                "Unclean leader election rate - increasing indicates broker failures",
                 "{election}"));
   }
 

jmx-metrics/src/main/resources/target-systems/kafka.groovy

@@ -167,29 +167,29 @@ otel.instrument(isrOperations,
 
 
 def maxLag = otel.mbean("kafka.server:type=ReplicaFetcherManager,name=MaxLag,clientId=Replica")
-otel.instrument(maxLag, "kafka.max.lag", "max lag in messages between follower and leader replicas",
+otel.instrument(maxLag, "kafka.max.lag", "Max lag in messages between follower and leader replicas",
         "{message}", "Value", otel.&longValueCallback)
 
 def activeControllerCount = otel.mbean("kafka.controller:type=KafkaController,name=ActiveControllerCount")
-otel.instrument(activeControllerCount, "kafka.controller.active.count", "controller is active on broker",
+otel.instrument(activeControllerCount, "kafka.controller.active.count", "For KRaft mode, the number of active controllers in the cluster. For ZooKeeper, indicates whether the broker is the controller broker.",
         "{controller}", "Value", otel.&longValueCallback)
 
 def leaderElectionRate = otel.mbean("kafka.controller:type=ControllerStats,name=LeaderElectionRateAndTimeMs")
-otel.instrument(leaderElectionRate, "kafka.leader.election.rate", "leader election rate - increasing indicates broker failures",
+otel.instrument(leaderElectionRate, "kafka.leader.election.rate", "Leader election rate - increasing indicates broker failures",
         "{election}", "Count", otel.&longCounterCallback)
 
 def uncleanLeaderElections = otel.mbean("kafka.controller:type=ControllerStats,name=UncleanLeaderElectionsPerSec")
-otel.instrument(uncleanLeaderElections, "kafka.unclean.election.rate", "unclean leader election rate - increasing indicates broker failures",
+otel.instrument(uncleanLeaderElections, "kafka.unclean.election.rate", "Unclean leader election rate - increasing indicates broker failures",
         "{election}", "Count", otel.&longCounterCallback)
 
 def requestQueueSize = otel.mbean("kafka.network:type=RequestChannel,name=RequestQueueSize")
-otel.instrument(requestQueueSize, "kafka.request.queue", "size of the request queue",
+otel.instrument(requestQueueSize, "kafka.request.queue", "Size of the request queue",
         "{request}", "Value", otel.&longValueCallback)
 
 def logFlushRate = otel.mbean("kafka.log:type=LogFlushStats,name=LogFlushRateAndTimeMs")
-otel.instrument(logFlushRate, "kafka.logs.flush.time.count", "log flush count",
+otel.instrument(logFlushRate, "kafka.logs.flush.time.count", "Log flush count",
         "ms", "Count", otel.&longCounterCallback)
-otel.instrument(logFlushRate, "kafka.logs.flush.time.median", "log flush time - 50th percentile",
+otel.instrument(logFlushRate, "kafka.logs.flush.time.median", "Log flush time - 50th percentile",
         "ms", "50thPercentile", otel.&doubleValueCallback)
-otel.instrument(logFlushRate, "kafka.logs.flush.time.99p", "log flush time - 99th percentile",
+otel.instrument(logFlushRate, "kafka.logs.flush.time.99p", "Log flush time - 99th percentile",
         "ms", "99thPercentile", otel.&doubleValueCallback)

jmx-scraper/src/integrationTest/java/io/opentelemetry/contrib/jmxscraper/target_systems/kafka/KafkaIntegrationTest.java

@@ -96,7 +96,7 @@ protected MetricsVerifier createMetricsVerifier() {
                 metric
                     .hasDescription("The number of requests waiting in purgatory")
                     .hasUnit("{request}")
-                    .isUpDownCounter()
+                    .isGauge()
                     .hasDataPointsWithAttributes(
                         attributeGroup(attribute("type", "Produce")),
                         attributeGroup(attribute("type", "Fetch"))))
@@ -140,7 +140,7 @@ protected MetricsVerifier createMetricsVerifier() {
                 metric
                     .hasDescription("Size of the request queue")
                     .hasUnit("{request}")
-                    .isUpDownCounter()
+                    .isGauge()
                     .hasDataPointsWithoutAttributes())
         .add(
             "kafka.partition.count",
@@ -180,20 +180,21 @@ protected MetricsVerifier createMetricsVerifier() {
             "kafka.controller.active.count",
             metric ->
                 metric
-                    .hasDescription("The number of controllers active on the broker") // CHANGED
+                    .hasDescription(
+                        "For KRaft mode, the number of active controllers in the cluster. For ZooKeeper, indicates whether the broker is the controller broker.")
                     .hasUnit("{controller}")
-                    .isUpDownCounter() // CHANGED
+                    .isGauge()
                     .hasDataPointsWithoutAttributes())
         .add(
-            "kafka.leader.election.count", // CHANGED from "kafka.leader.election.rate"
+            "kafka.leader.election.rate",
             metric ->
                 metric
                     .hasDescription("The leader election count")
                     .hasUnit("{election}")
                     .isCounter()
                     .hasDataPointsWithoutAttributes())
         .add(
-            "kafka.leader.election.unclean.count", // CHANGED from "kafka.unclean.election.rate"
+            "kafka.unclean.election.rate",
             metric ->
                 metric
                     .hasDescription(
@@ -202,40 +203,13 @@ protected MetricsVerifier createMetricsVerifier() {
                     .isCounter()
                     .hasDataPointsWithoutAttributes())
         .add(
-            "kafka.lag.max", // CHANGED from "kafka.max.lag"
+            "kafka.max.lag",
             metric ->
                 metric
                     .hasDescription(
                         "The max lag in messages between follower and leader replicas") // CHANGED
                     .hasUnit("{message}")
                     .isGauge()
-                    .hasDataPointsWithoutAttributes())
-
-    // TODO: Find out how to force Kafka to generate these metrics
-    //        .add(
-    //            "kafka.logs.flush.count",
-    //            metric ->
-    //                metric
-    //                    .hasDescription("Log flush count")
-    //                    .hasUnit("{flush}")
-    //                    .isCounter()
-    //                    .hasDataPointsWithoutAttributes())
-    //        .add(
-    //            "kafka.logs.flush.time.median",
-    //            metric ->
-    //                metric
-    //                    .hasDescription("Log flush time - 50th percentile")
-    //                    .hasUnit("ms")
-    //                    .isGauge()
-    //                    .hasDataPointsWithoutAttributes())
-    //        .add(
-    //            "kafka.logs.flush.time.99p",
-    //            metric ->
-    //                metric
-    //                    .hasDescription("Log flush time - 99th percentile")
-    //                    .hasUnit("ms")
-    //                    .isGauge()
-    //                    .hasDataPointsWithoutAttributes())
-    ;
+                    .hasDataPointsWithoutAttributes());
   }
 }

jmx-scraper/src/main/resources/kafka.yaml

@@ -80,7 +80,7 @@ rules:
     mapping:
       Value:
         metric: kafka.request.queue
-        type: updowncounter
+        type: gauge
         desc: Size of the request queue
         unit: "{request}"
 
@@ -112,7 +112,7 @@ rules:
     mapping:
       Value:
         metric: kafka.purgatory.size
-        type: updowncounter
+        type: gauge
         desc: The number of requests waiting in purgatory
         unit: "{request}"
 
@@ -163,30 +163,32 @@ rules:
   - bean: kafka.server:type=ReplicaFetcherManager,name=MaxLag,clientId=Replica
     mapping:
       Value:
-        metric: kafka.lag.max
+        metric: kafka.max.lag
+        type: gauge
         desc: The max lag in messages between follower and leader replicas
         unit: "{message}"
 
+
   - bean: kafka.controller:type=KafkaController,name=ActiveControllerCount
     mapping:
       Value:
         metric: kafka.controller.active.count
-        type: updowncounter
-        desc: The number of controllers active on the broker
+        type: gauge
+        desc: For KRaft mode, the number of active controllers in the cluster. For ZooKeeper, indicates whether the broker is the controller broker.
         unit: "{controller}"
 
   - bean: kafka.controller:type=ControllerStats,name=LeaderElectionRateAndTimeMs
     mapping:
       Count:
-        metric: kafka.leader.election.count
+        metric: kafka.leader.election.rate
         type: counter
         desc: The leader election count
         unit: "{election}"
 
   - bean: kafka.controller:type=ControllerStats,name=UncleanLeaderElectionsPerSec
     mapping:
       Count:
-        metric: kafka.leader.election.unclean.count
+        metric: kafka.unclean.election.rate
         type: counter
         desc: Unclean leader election count - increasing indicates broker failures
         unit: "{election}"