I'm new to kafka and am setting up an ELK stack to collect JMX metrics and log files for analysis and reporting.
Currently, I've setup my logstash to collect kafka mbeans in the following patterns:
{
"object_name" : "kafka.*:type=*,name=*,topic=*,partition=*"
},
{
"object_name" : "kafka.*:type=*,name=*,topic=*"
},
{
"object_name" : "kafka.*:type=*,name=*"
},
{
"object_name" : "kafka.*:type=*,client-id=*"
},
{
"object_name" : "kafka.*:type=*,name=*,clientId=*,brokerHost=*,brokerPort=*"
},
{
"object_name" : "kafka.*:type=*,name=*,clientId=*"
},
{
"object_name" : "kafka.*:type=*,broker-id=*,fetcher-id=*"
},
{
"object_name" : "kafka.*:type=*,client-id=*"
}
The issue that I'm having is that I'm missing all kafka.consumer
metrics and when I log into to jconsole on the RMI port, I don't see kafka.consumer
mbeans there either.
I'm new to this whole JMX/mbeans stuff and kafka, so am I looking in the right place for kafka.consumer
metrics? I currently do have active producers and consumers hitting the cluster.