Hello all,
I'm trying to set up some monitoring of a confluentinc/cp-server-connect:7.6.0
docker container and the connectors I run within that. I am using the input-jmx plugin.
Now I would like to separate the pipelines for every connector as they have different attributes/metrics.
My pipelines.yml
looks like this:
- pipeline.id: ODataV2Source
path.config: "/etc/logstash/pipeline/odatav2source.conf"
odatav2source.conf
input part looks like:
# ==================
# 1) INPUT
# ==================
input {
jmx {
path => "/usr/share/logstash/jmxconf"
polling_frequency => 15
nb_thread => 4
type => "jmx"
}
}
and my odatav2source.json
in /usr/share/logstash/jmxconf
:
{
"host": "connect",
"port": 9999,
"alias": "kafka.connect",
"queries": [
{
"object_name": "org.init.ohja.kafka.connect:connector=*,task=*,type=odatav2-source-task-metrics",
"attributes": [
"0-entityset",
"0-position",
"0-service",
"0-service-url",
"0-topic",
"0-uri-type",
"active-subscriptions",
"last-extraction",
"retries"
],
"object_alias": "${type}.${connector}.${task}"
},...
now I would like to have a .conf
and .json
file for every connector and also possibly put every connectors metrics in a separate index to keep thing clean and separated.
However setting the path in the jmx input to "/usr/share/logstash/jmxconf/odatav2source.json"
does not work and results in this error:
[ODataV2Source][many numbers] Not a directory - /usr/share/logstash/jmxconf/odatav2source.json
Any help in setting this up would be appreciated!
Best regards
David