config.yml

# Note: we add the environment label to all of our metrics, but if you don’t have a similar label, you’ll need to remove all references to it
- name: kafka_sla_rules
  rules:
  - alert: kafka_sla_produce_availability
    expr: kafka_monitor_produce_service_produce_availability_avg_1m{environment!="loadtest"} < 1
    for: 5m
    labels:
      severity: page
    annotations:
      description: '{{ $labels.environment }} kafka produce availability is {{ $value }}'
      summary: '{{ $labels.environment }} kafka produce availability is < 1 for more than 5 minutes'

  - alert: kafka_sla_consume_availability
    expr: kafka_monitor_consume_service_consume_availability_avg_1m{environment!="loadtest"} < 1
    for: 5m
    labels:
      severity: page
    annotations:
      description: '{{ $labels.environment }} kafka consume availability is {{ $value }}'
      summary: '{{ $labels.environment }} kafka consume availability is < 1 for more than 5 minutes'


  - alert: kafka_under_replicated_partitions
    expr: sum(kafka_server_replicamanager_value{name="UnderReplicatedPartitions",environment!="loadtest"}) by(environment) > 0
    for: 5m
    labels:
      severity: page
    annotations:
      description: '{{ $labels.environment }} kafka is reporting {{ $value }}
        under-replicated partitions.'
      summary: '{{ $labels.environment }} kafka under-replicated partitions'

  - alert: kafka_offline_partitions
    expr: sum(kafka_controller_kafkacontroller_value{name="OfflinePartitionsCount",environment!="loadtest"}) by(environment) > 0
    for: 5m
    labels:
      severity: page
    annotations:
      description: '{{ $labels.environment }} kafka is reporting {{ $value }} offline
        partitions. This will prevent anything from producing or consuming from that partition.'
      summary: '{{ $labels.environment }} kafka offline partitions'

  - alert: kafka_no_active_controller
    expr: sum(kafka_controller_kafkacontroller_value{job="kafka", name="ActiveControllerCount",environment!="loadtest"}) by(environment) != 1
    for: 5m
    labels:
      severity: page
    annotations:
      description: '{{ $labels.environment }} kafka is reporting no active controllers. Restarting one broker may
        resolve this.'
      summary: '{{ $labels.environment }} kafka has no active controller'