Package com.linkedin.davinci.stats

Class KafkaConsumerServiceStats

java.lang.Object

com.linkedin.venice.stats.AbstractVeniceStats

com.linkedin.davinci.stats.KafkaConsumerServiceStats

public class KafkaConsumerServiceStats
extends AbstractVeniceStats

This class provides the stats for Kafka consumer service per region or per store. Stats inside this class can either: (1) Total only: The stat indicate total number of all the stores on this host per region. (2) Total and Per store only: The stat is registered for each store on this host.

Field Summary

Fields inherited from class com.linkedin.venice.stats.AbstractVeniceStats

DELIMITER

Constructor Summary

Constructors

Constructor	Description
KafkaConsumerServiceStats(io.tehuti.metrics.MetricsRepository metricsRepository, String storeName, LongSupplier getMaxElapsedTimeSinceLastPollInConsumerPool, KafkaConsumerServiceStats totalStats, Time time, String clusterName)

Method Summary

Modifier and Type	Method	Description
void	recordAvgPartitionsPerConsumer(int count)
void	recordByteSizePerPoll(double count)	Joint API — records to both Tehuti (Min, Max) and OTel (MIN_MAX_COUNT_SUM).
void	recordConsumerIdleTime(double time)	Joint API — records to both Tehuti (Max) and OTel (MIN_MAX_COUNT_SUM).
void	recordConsumerRecordsProducingToWriterBufferLatency(double latency)	Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM).
void	recordDelegateSubscribeLatency(double value)	Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) with SUBSCRIBE dimension.
void	recordDetectedDeletedTopicNum(int count)	Joint API — records to both Tehuti (Total) and OTel (COUNTER).
void	recordDetectedNoRunningIngestionTopicPartitionNum(int count)	Joint API — records to both Tehuti (Total) and OTel (COUNTER).
void	recordMaxPartitionsPerConsumer(int count)
void	recordMinPartitionsPerConsumer(int count)
void	recordNonZeroPollResultNum(int count)	Joint API — records to both Tehuti (LongAdderRateGauge) and OTel (ASYNC_COUNTER_FOR_HIGH_PERF_CASES).
void	recordPartitionAssignmentForOtel(int partitionCount)	Records a single per-consumer partition count to the OTel partition assignment histogram.
void	recordPollError()	Joint API — records to both Tehuti (OccurrenceRate) and OTel (COUNTER).
void	recordPollRequestLatency(double latency)	Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) for poll time, plus poll count.
void	recordPollResultNum(int count)	Joint API — records to both Tehuti (Avg, Min) and OTel (MIN_MAX_COUNT_SUM).
void	recordSubscribedPartitionsNum(int count)
void	recordUpdateCurrentAssignmentLatency(double value)	Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) with UPDATE_ASSIGNMENT dimension.

Methods inherited from class com.linkedin.venice.stats.AbstractVeniceStats

avgAndMax, avgAndTotal, getMetricFullName, getMetricsRepository, getName, getSensorFullName, getSensorFullName, isTotalStats, minAndMax, registerOnlyTotalRate, registerOnlyTotalSensor, registerPerStoreAndTotal, registerPerStoreAndTotalSensor, registerSensor, registerSensor, registerSensor, registerSensor, registerSensorAttributeGauge, registerSensorIfAbsent, registerSensorIfAbsent, registerSensorIfAbsent, registerSensorWithAggregate, registerSensorWithAggregate, unregisterAllSensors

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

KafkaConsumerServiceStats

public KafkaConsumerServiceStats(io.tehuti.metrics.MetricsRepository metricsRepository,
 String storeName,
 LongSupplier getMaxElapsedTimeSinceLastPollInConsumerPool,
 KafkaConsumerServiceStats totalStats,
 Time time,
 String clusterName)

Method Details

recordPollRequestLatency

public void recordPollRequestLatency(double latency)

Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) for poll time, plus poll count.

recordPollResultNum

public void recordPollResultNum(int count)

Joint API — records to both Tehuti (Avg, Min) and OTel (MIN_MAX_COUNT_SUM).

recordNonZeroPollResultNum

public void recordNonZeroPollResultNum(int count)

Joint API — records to both Tehuti (LongAdderRateGauge) and OTel (ASYNC_COUNTER_FOR_HIGH_PERF_CASES).

recordConsumerRecordsProducingToWriterBufferLatency

public void recordConsumerRecordsProducingToWriterBufferLatency(double latency)

Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM).

recordPollError

public void recordPollError()

Joint API — records to both Tehuti (OccurrenceRate) and OTel (COUNTER).

recordDetectedDeletedTopicNum

public void recordDetectedDeletedTopicNum(int count)

Joint API — records to both Tehuti (Total) and OTel (COUNTER).

recordDetectedNoRunningIngestionTopicPartitionNum

public void recordDetectedNoRunningIngestionTopicPartitionNum(int count)

Joint API — records to both Tehuti (Total) and OTel (COUNTER).

recordDelegateSubscribeLatency

public void recordDelegateSubscribeLatency(double value)

Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) with SUBSCRIBE dimension.

recordUpdateCurrentAssignmentLatency

public void recordUpdateCurrentAssignmentLatency(double value)

Joint API — records to both Tehuti (Avg, Max) and OTel (HISTOGRAM) with UPDATE_ASSIGNMENT dimension.

recordMinPartitionsPerConsumer

public void recordMinPartitionsPerConsumer(int count)

recordMaxPartitionsPerConsumer

public void recordMaxPartitionsPerConsumer(int count)

recordAvgPartitionsPerConsumer

public void recordAvgPartitionsPerConsumer(int count)

recordByteSizePerPoll

public void recordByteSizePerPoll(double count)

Joint API — records to both Tehuti (Min, Max) and OTel (MIN_MAX_COUNT_SUM).

recordConsumerIdleTime

public void recordConsumerIdleTime(double time)

Joint API — records to both Tehuti (Max) and OTel (MIN_MAX_COUNT_SUM).

recordSubscribedPartitionsNum

public void recordSubscribedPartitionsNum(int count)

recordPartitionAssignmentForOtel

public void recordPartitionAssignmentForOtel(int partitionCount)

Records a single per-consumer partition count to the OTel partition assignment histogram. Called for each consumer in the pool from KafkaConsumerService.recordPartitionsPerConsumerSensor(). Tehuti keeps 4 pre-computed gauge values; OTel records raw per-consumer values to a single MIN_MAX_COUNT_SUM histogram.