kafka_exporter

Kafka exporter for Prometheus. For other metrics from Kafka, have a look at the JMX exporter.

Table of Contents

• Compatibility
• Dependency
• Download
• Compile
  • Build Binary
  • Build Docker Image
• Run
  • Run Binary
  • Run Docker Image
• Flags
  • Notes
• Metrics
  • Brokers
  • Topics
  • Consumer Groups
• Grafana Dashboard
• Contribute
• Donation
• License

Compatibility

Support Apache Kafka version 0.10.1.0 (and later).

Dependency

• Prometheus
• Sarama
• Golang

Download

Binary can be downloaded from Releases page.

Compile

Build Binary

make

Build Docker Image

make docker

Docker Hub Image

docker pull dparrott/kafka-exporter:latest

It can be used directly instead of having to build the image yourself Docker Hub dparrott/kafka-exporter

Run

Run Binary

kafka_exporter --kafka.server=kafka:9092 [--kafka.server=another-server ...]

Run Docker Image

docker run -ti --rm -p 9308:9308 dparrott/kafka-exporter --kafka.server=kafka:9092 [--kafka.server=another-server ...]

Flags

This image is configurable using different flags

Flag name	Default	Description
kafka.server	kafka:9092	Addresses (host:port) of Kafka server
kafka.version	1.0.0	Kafka broker version
sasl.enabled	false	Connect using SASL/PLAIN
sasl.handshake	true	Only set this to false if using a non-Kafka SASL proxy
sasl.username		SASL user name
sasl.password		SASL user password
sasl.mechanism		SASL mechanism can be plain, scram-sha512, scram-sha256
tls.enabled	false	Connect using TLS
tls.ca-file		The optional certificate authority file for TLS client authentication
tls.cert-file		The optional certificate file for client authentication
tls.key-file		The optional key file for client authentication
tls.insecure-skip-tls-verify	false	If true, the server's certificate will not be checked for validity
topic.filter	.*	Regex that determines which topics to collect
group.filter	.*	Regex that determines which consumer groups to collect
web.listen-address	:9308	Address to listen on for web interface and telemetry
web.telemetry-path	/metrics	Path under which to expose metrics
log.level	info	Only log messages with the given severity or above. Valid levels: [debug, info, warn, error, fatal]
log.enable-sarama	false	Turn on Sarama logging
max.offsets	1000	Maximum number of offsets to store in the interpolation table for a partition
prune.interval	30	How frequently should the interpolation table be pruned, in seconds

Notes

Boolean values are uniquely managed by Kingpin. Each boolean flag will have a negative complement: --<name> and --no-<name>.

For example:

If you need to disable sasl.handshake, you could add flag --no-sasl.handshake

Metrics

Documents about exposed Prometheus metrics.

For details on the underlying metrics please see Apache Kafka.

Brokers

Metrics details

Name	Exposed information
kafka_brokers	Number of Brokers in the Kafka Cluster

Metrics output example

# HELP kafka_brokers Number of Brokers in the Kafka Cluster.
# TYPE kafka_brokers gauge
kafka_brokers 3

Topics

Metrics details

Name	Exposed information
kafka_topic_partitions	Number of partitions for this Topic
kafka_topic_partition_current_offset	Current Offset of a Broker at Topic/Partition
kafka_topic_partition_oldest_offset	Oldest Offset of a Broker at Topic/Partition
kafka_topic_partition_in_sync_replica	Number of In-Sync Replicas for this Topic/Partition
kafka_topic_partition_leader	Leader Broker ID of this Topic/Partition
kafka_topic_partition_leader_is_preferred	1 if Topic/Partition is using the Preferred Broker
kafka_topic_partition_replicas	Number of Replicas for this Topic/Partition
kafka_topic_partition_under_replicated_partition	1 if Topic/Partition is under Replicated

Metrics output example

# HELP kafka_topic_partitions Number of partitions for this Topic
# TYPE kafka_topic_partitions gauge
kafka_topic_partitions{topic="__consumer_offsets"} 50

# HELP kafka_topic_partition_current_offset Current Offset of a Broker at Topic/Partition
# TYPE kafka_topic_partition_current_offset gauge
kafka_topic_partition_current_offset{partition="0",topic="__consumer_offsets"} 0

# HELP kafka_topic_partition_oldest_offset Oldest Offset of a Broker at Topic/Partition
# TYPE kafka_topic_partition_oldest_offset gauge
kafka_topic_partition_oldest_offset{partition="0",topic="__consumer_offsets"} 0

# HELP kafka_topic_partition_in_sync_replica Number of In-Sync Replicas for this Topic/Partition
# TYPE kafka_topic_partition_in_sync_replica gauge
kafka_topic_partition_in_sync_replica{partition="0",topic="__consumer_offsets"} 3

# HELP kafka_topic_partition_leader Leader Broker ID of this Topic/Partition
# TYPE kafka_topic_partition_leader gauge
kafka_topic_partition_leader{partition="0",topic="__consumer_offsets"} 0

# HELP kafka_topic_partition_leader_is_preferred 1 if Topic/Partition is using the Preferred Broker
# TYPE kafka_topic_partition_leader_is_preferred gauge
kafka_topic_partition_leader_is_preferred{partition="0",topic="__consumer_offsets"} 1

# HELP kafka_topic_partition_replicas Number of Replicas for this Topic/Partition
# TYPE kafka_topic_partition_replicas gauge
kafka_topic_partition_replicas{partition="0",topic="__consumer_offsets"} 3

# HELP kafka_topic_partition_under_replicated_partition 1 if Topic/Partition is under Replicated
# TYPE kafka_topic_partition_under_replicated_partition gauge
kafka_topic_partition_under_replicated_partition{partition="0",topic="__consumer_offsets"} 0

Consumer Groups

Metrics details

Name	Exposed information
kafka_consumergroup_current_offset	Current Offset of a ConsumerGroup at Topic/Partition
kafka_consumergroup_lag	Current Approximate Lag of a ConsumerGroup at Topic/Partition

Metrics output example

# HELP kafka_consumergroup_current_offset Current Offset of a ConsumerGroup at Topic/Partition
# TYPE kafka_consumergroup_current_offset gauge
kafka_consumergroup_current_offset{consumergroup="KMOffsetCache-kafka-manager-3806276532-ml44w",partition="0",topic="__consumer_offsets"} -1

# HELP kafka_consumergroup_lag Current Approximate Lag of a ConsumerGroup at Topic/Partition
# TYPE kafka_consumergroup_lag gauge
kafka_consumergroup_lag{consumergroup="KMOffsetCache-kafka-manager-3806276532-ml44w",partition="0",topic="__consumer_offsets"} 1

Consumer Lag

Metric Details

Name	Exposed information
kafka_consumer_lag_millis	Current approximation of consumer lag for a ConsumerGroup at Topic/Partition
kafka_consumer_lag_extrapolation	Indicates that a consumer group lag estimation used extrapolation
kafka_consumer_lag_interpolation	Indicates that a consumer group lag estimation used interpolation

Metrics output example

# HELP kafka_consumer_lag_extrapolation Indicates that a consumer group lag estimation used extrapolation
# TYPE kafka_consumer_lag_extrapolation counter
kafka_consumer_lag_extrapolation{consumergroup="perf-consumer-74084",partition="0",topic="test"} 1
   
# HELP kafka_consumer_lag_interpolation Indicates that a consumer group lag estimation used interpolation
# TYPE kafka_consumer_lag_interpolation counter
kafka_consumer_lag_interpolation{consumergroup="perf-consumer-74084",partition="0",topic="test"} 1
   
# HELP kafka_consumer_lag_millis Current approximation of consumer lag for a ConsumerGroup at Topic/Partition
# TYPE kafka_consumer_lag_millis gauge
kafka_consumer_lag_millis{consumergroup="perf-consumer-74084",partition="0",topic="test"} 3.4457231197552e+10

Grafana Dashboard

Grafana Dashboard ID: 7589, name: Kafka Exporter Overview.

For details of the dashboard please see Kafka Exporter Overview.

Lag Estimation

The technique to estimate lag for a consumer group, topic, and partition is taken from the Lightbend Kafka Lag Exporter.

Once the exporter starts up, sampling of the next offset to be produced begins. The interpolation table is built from these samples, and the current offset for each monitored consumer group are compared against values in the table. If an upper and lower bound for the current offset of a consumer group are in the table, the interpolation technique is used. If only an upper bound is container within the table, extrapolation is used.

At a configurable interval prune.interval (default is 30 seconds) an operation to prune the interpolation table is performed. Any consumer group or topic that are no longer listed by the broker is removed. The number of offsets for each partition is trimmed down to max.offsets (default 1000), with the oldest offsets removed first.

Pruning of the interpolation table happens on a separate thread and thread safety is ensured by a lock around the interpolation table.

Contribute

To contribute to the upstream project, please open a pull request.

To contribute to this fork please open a pull request here

Donation

To donate to the developer of the project this is forked from please use the donation link below

License

Code is licensed under the Apache License 2.0.