Details
-
Bug
-
Status: Open
-
Critical
-
Resolution: Unresolved
-
3.5.8
-
None
-
None
Description
Deploy kafka and zookeeper in k8s, and use the following configuration to output monitoring metrics
-javaagent:/opt/kafka/libs/jmx_prometheus_javaagent-0.12.0.jar=9404:/opt/kafka/custom-config/metrics-config.yml
The metrics configuration is as follows:
$ cat /opt/kafka/custom-config/metrics-config.yml
{
"lowercaseOutputName":true,
"rules":[
{ "name":"zookeeper_$2", "pattern":"org.apache.ZooKeeperService<name0=ReplicatedServer_id(\\d+)><>(\\w+)", "type":"GAUGE" }
,
{
"labels":
{ "replicaId":"$2" }
,
"name":"zookeeper_$3",
"pattern":"org.apache.ZooKeeperService<name0=ReplicatedServer_id(d+), name1=replica.(\\d+)><>(w+)",
"type":"GAUGE"
},
{
"labels":
{ "memberType":"$3", "replicaId":"$2" }
,
"name":"zookeeper_$4",
"pattern":"org.apache.ZooKeeperService<name0=ReplicatedServer_id(d+), name1=replica.(d+), name2=(w+)><>(Packets.*)",
"type":"COUNTER"
},
{
"labels":
{ "memberType":"$3", "replicaId":"$2" }
,
"name":"zookeeper_$4",
"pattern":"org.apache.ZooKeeperService<name0=ReplicatedServer_id(d+), name1=replica.(d+), name2=(\\w+)><>(w+)",
"type":"GAUGE"
},
{
"labels":
{ "memberType":"$3", "replica":"$2" }
,
"name":"zookeeper_$5",
"pattern":"org.apache.ZooKeeperService<name0=(.).name1=(.).name2=(.).name3=(.)><>(PacketsReceived)",
"type":"COUNTER"
},
{
"labels":
{ "memberType":"$3", "replica":"$2" }
,
"name":"zookeeper_$5",
"pattern":"org.apache.ZooKeeperService<name0=(.).name1=(.).name2=(.).name3=(.)><>(PacketsSent)",
"type":"COUNTER"
},
{
"labels":
{ "memberType":"$3", "replicaId":"$2" }
,
"name":"zookeeper_$4_$5",
"pattern":"org.apache.ZooKeeperService<name0=ReplicatedServer_id(d+), name1=replica.(d+), name2=(w+), name3=(\\w+)><>(w+)",
"type":"GAUGE"
},
{ "name":"zookeeper_$2", "pattern":"org.apache.ZooKeeperService<name0=StandaloneServer_port(\\d+)><>(\\w+)", "type":"GAUGE" }
,
{ "name":"zookeeper_$2", "pattern":"org.apache.ZooKeeperService<name0=StandaloneServer_port(\\d+), name1=InMemoryDataTree><>(\\w+)", "type":"GAUGE" }
]
}
At this time, there are some very strange monitoring metrics in promethues
The jmx exporter version used is 0.12.0
What is the problem at this time, have you encountered it?
