apoorvmittal10 commented on code in PR #14620: URL: https://github.com/apache/kafka/pull/14620#discussion_r1390693105
########## clients/src/main/java/org/apache/kafka/common/telemetry/internals/KafkaMetricsCollector.java: ########## @@ -0,0 +1,341 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.kafka.common.telemetry.internals; + +import org.apache.kafka.common.KafkaException; +import org.apache.kafka.common.MetricName; +import org.apache.kafka.common.metrics.Gauge; +import org.apache.kafka.common.metrics.KafkaMetric; +import org.apache.kafka.common.metrics.Measurable; +import org.apache.kafka.common.metrics.MetricValueProvider; +import org.apache.kafka.common.metrics.stats.Avg; +import org.apache.kafka.common.metrics.stats.CumulativeCount; +import org.apache.kafka.common.metrics.stats.CumulativeSum; +import org.apache.kafka.common.metrics.stats.Frequencies; +import org.apache.kafka.common.metrics.stats.Max; +import org.apache.kafka.common.metrics.stats.Meter; +import org.apache.kafka.common.metrics.stats.Min; +import org.apache.kafka.common.metrics.stats.Percentiles; +import org.apache.kafka.common.metrics.stats.Rate; +import org.apache.kafka.common.metrics.stats.SimpleRate; +import org.apache.kafka.common.metrics.stats.WindowedCount; +import org.apache.kafka.common.telemetry.internals.LastValueTracker.InstantAndValue; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import java.lang.reflect.Field; +import java.time.Clock; +import java.time.Instant; +import java.util.List; +import java.util.Map; +import java.util.Map.Entry; +import java.util.Optional; +import java.util.Set; +import java.util.concurrent.ConcurrentHashMap; + +/** + * All metrics implement the {@link MetricValueProvider} interface. They are divided into + * two base types: + * + * <ol> + * <li>{@link Gauge}</li> + * <li>{@link Measurable}</li> + * </ol> + * + * {@link Gauge Gauges} can have any value but we only collect metrics with number values. + * {@link Measurable Measurables} are divided into simple types with single values + * ({@link Avg}, {@link CumulativeCount}, {@link Min}, {@link Max}, {@link Rate}, + * {@link SimpleRate}, and {@link CumulativeSum}) and compound types ({@link Frequencies}, + * {@link Meter}, and {@link Percentiles}). + * + * <p> + * + * We can safely assume that a {@link CumulativeCount count} always increases in steady state. It + * should be a bug if a count metric decreases. + * + * <p> + * + * Total and Sum are treated as a monotonically increasing counter. The javadocs for Total metric type + * say "An un-windowed cumulative total maintained over all time.". The standalone Total metrics in + * the codebase seem to be cumulative metrics that will always increase. The Total metric underlying + * Meter type is mostly a Total of a Count metric. + * We can assume that a Total metric always increases (but it is not guaranteed as the sample values might be both + * negative or positive). + * For now, Total is converted to CUMULATIVE_DOUBLE unless we find a valid counter-example. + * + * <p> + * + * The Sum as it is a sample sum which is not a cumulative metric. It is converted to GAUGE_DOUBLE. + * + * <p> + * + * The compound metrics are virtual metrics. They are composed of simple types or anonymous measurable types + * which are reported. A compound metric is never reported as-is. + * + * <p> + * + * A Meter metric is always created with and reported as 2 KafkaExporter metrics: a rate and a + * count. For eg: org.apache.kafka.common.network.Selector has Meter metric for "connection-close" but it + * has to be created with a "connection-close-rate" metric of type rate and a "connection-close-total" + * metric of type total. So, we will never get a KafkaExporter metric with type Meter. + * + * <p> + * + * Frequencies is created with a array of Frequency objects. When a Frequencies metric is registered, each + * member Frequency object is converted into an anonymous Measurable and registered. So, a Frequencies metric + * is reported with a set of measurables with name = Frequency.name(). As there is no way to figure out the + * compound type, each component measurables is converted to a GAUGE_DOUBLE. + * + * <p> + * + * Percentiles work the same way as Frequencies. The only difference is that it is composed of Percentile + * types instead. So, we should treat the component measurable as GAUGE_DOUBLE. + * + * <p> + * + * Some metrics are defined as either anonymous inner classes or lambdas implementing the Measurable + * interface. As we do not have any information on how to treat them, we should fallback to treating + * them as GAUGE_DOUBLE. + * + * <p> + * + * KafkaExporter -> OpenTelemetry mapping for measurables Review Comment: My bad, corrected the comment. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
