Group Coordinator Metrics

This topic describes JMX metrics for the Kafka group coordinator. These metrics are useful for monitoring consumer group management in your Kafka cluster.

For information about how to configure JMX, see Configure JMX for Monitoring.

Search for a metric

The following metrics are available for monitoring the group coordinator.

batch-flush-time-ms

MBean: kafka.server:type=group-coordinator-metrics,name=batch-flush-time-ms-[max|p50|p95|p99|p999]: The time that a batch took to be flushed to the local partition.

consumer-group-count

MBean: kafka.server:type=group-coordinator-metrics,name=consumer-group-count,state=[empty|assigning|reconciling|stable|dead]: Total number of consumer groups in each state. Possible states can be Empty, Assigning, Reconciling, Stable, or Dead.

consumer-group-rebalance-count

MBean: kafka.server:type=group-coordinator-metrics,name=consumer-group-rebalance-count: Total number of consumer group rebalances.

consumer-group-rebalance-rate

MBean: kafka.server:type=group-coordinator-metrics,name=consumer-group-rebalance-rate: The rebalance rate of consumer groups.

event-processing-time-ms

MBean: kafka.server:type=group-coordinator-metrics,name=event-processing-time-ms-[max|p50|p95|p99|p999]: The time that an event took to be processed.

event-purgatory-time-ms

MBean: kafka.server:type=group-coordinator-metrics,name=event-purgatory-time-ms-[max|p50|p95|p99|p999]: The time that an event waited in the purgatory before being completed.

event-queue-size

MBean: kafka.server:type=group-coordinator-metrics,name=event-queue-size: The number of events waiting to be processed in the queue.

event-queue-time-ms

MBean: kafka.server:type=group-coordinator-metrics,name=event-queue-time-ms-[max|p50|p95|p99|p999]: The time that an event spent waiting in the queue to be processed.

group-completed-rebalance-count

MBean: kafka.server:type=group-coordinator-metrics,name=group-completed-rebalance-count: The total number of classic group completed rebalances.

group-completed-rebalance-rate

MBean: kafka.server:type=group-coordinator-metrics,name=group-completed-rebalance-rate: The rate of classic group completed rebalances.

group-count

MBean: kafka.server:type=group-coordinator-metrics,name=group-count,protocol={consumer|classic}: Total number of groups per group type. Possible types can be Classic or Consumer.

NumGroups (Classic)

MBean: kafka.server:type=GroupMetadataManager,name=NumGroups: Total number of Classic Groups.

NumGroups by state (Classic)

MBean: kafka.server:type=GroupMetadataManager,name=NumGroups[PreparingRebalance,CompletingRebalance,Empty,Stable,Dead]: The number of Classic Groups in each state. Possible states can be PreparingRebalance, CompletingRebalance, Empty, Stable, or Dead.

NumOffsets

MBean: kafka.server:type=GroupMetadataManager,name=NumOffsets: Total number of committed offsets for classic and consumer groups.

offset-commit-count

MBean: kafka.server:type=group-coordinator-metrics,name=offset-commit-count: The total number of committed offsets.

offset-commit-rate

MBean: kafka.server:type=group-coordinator-metrics,name=offset-commit-rate: The rate of committed offsets.

offset-deletion-count

MBean: kafka.server:type=group-coordinator-metrics,name=offset-deletion-count: The total number of administrative deleted offsets.

offset-deletion-rate

MBean: kafka.server:type=group-coordinator-metrics,name=offset-deletion-rate: The rate of administrative deleted offsets.

offset-expiration-count

MBean: kafka.server:type=group-coordinator-metrics,name=offset-expiration-count: The total number of expired offsets.

offset-expiration-rate

MBean: kafka.server:type=group-coordinator-metrics,name=offset-expiration-rate: The rate of expired offsets.

Partition count by state

MBean: kafka.server:type=group-coordinator-metrics,name=num-partitions,state={loading|active|failed}: The number of __consumer_offsets partitions hosted by the broker, broken down by state. Possible states can be loading, active, or failed.

partition-load-time-avg

MBean: kafka.server:type=group-coordinator-metrics,name=partition-load-time-avg: The average loading time needed to read the state from the __consumer_offsets partitions.

partition-load-time-max

MBean: kafka.server:type=group-coordinator-metrics,name=partition-load-time-max: The maximum loading time needed to read the state from the __consumer_offsets partitions.

thread-idle-ratio-avg

MBean: kafka.server:type=group-coordinator-metrics,name=thread-idle-ratio-avg: The average idle ratio of the coordinator threads.