App Configuration

The most flexible way to start Kafka Monitor is to run kafka-monitor-start.sh with a config file, which allows you to instantiate multiple Service or App that are already implemented in Kafka Monitor and tune their configs to monitor your clusters. See here for an example config file.

In this section we list the configurations of the App classes that have been implemented in Kafka Monitor.

SingleClusterMonitor Configuration Parameters

SingleClusterMonitor directly passes configs to ProduceService and ConsumeService. Thus its configuration parameters are the union of the configuration parameters of these two services. See Service Configuration for parameters of ProduceService and ConsumeService.

Configure ProduceService and ConsumeService in Different JVMs

Here's an example config that just runs the ConsumeService. Please include produce-service in combination with the following for the metrics to work:

"consume-service": {
    "class.name": "com.linkedin.kmf.services.ConsumeService",
    "topic": "kafka-monitor-topic",
    "zookeeper.connect": "localhost:2181",
    "bootstrap.servers": "localhost:9092",
    "consume.latency.sla.ms": "20000",
    "consume.consumer.props": {
    }
  }