Data not coming out on JMX integration

I have problem getting JMX infos into New relic from K8s pods. I’m using the image newrelic/infrastructure-k8s:2.1.0, and here is my configuration file.

apiVersion: v1
kind: ConfigMap
  name: nri-integration-cfg # aimed to be safely overridden by users
  namespace: default
  jmx-config.yml: |
        exec: /var/db/newrelic-infra/nri-discovery-kubernetes
      - name: nri-jmx
          JMX_HOST: ${discovery.ip}
          JMX_PORT: 9999
          COLLECTION_FILES: ${config.path}
            - domain: org.apache.flink.*
              event_type: FlinkMetricSample
                - query: host = *
          env: staging

I tried the jmxterm tool, it is ok to connect to the destination pod with <clusterIP>:9999 and see the MBeans. But the integration just don’t work, nothing is comming out as FlinkMetricSamples.

And I opened verbose log, only the following lines about jmx comes out. No error is seen.

time="2021-01-13T09:24:20Z" level=debug msg="Running task." component=integrations.Definition integration_name=nri-jmx
time="2021-01-13T09:24:20Z" level=debug msg="Running through all discovery matches." component=integrations.Definition integration_name=nri-jmx
time="2021-01-13T09:24:20Z" level=debug msg="Waiting while the integration instances run." component=integrations.runner.Runner env=staging integration_name=nri-jmx
time="2021-01-13T09:24:20Z" level=debug msg="Integration instances finished their execution. Waiting until next interval." component=integrations.runner.Runner env=staging integration_name=nri-jmx

I’m wondering which stage goes wrong, is the discovery or connect part? Can anyone help with this?

Hi @lyjlulyj, Sorry you have been waiting awhile for a response from our community.
I’m going to bring this back to the attention of our support team. Thanks for your patience!

Neal Mc

Did anyone get an answer to this? Im seeing same issue

Hi is this on supports radar? Im seeing the same issue with newrelic/infrastructure-k8s:2.2.0, and also wondering if its the discovery is failing, the debug tracing could be better

time=“2021-02-03T12:02:08Z” level=debug msg=“Integration definition loaded.” component=PluginRegistry integration=com.newrelic.jmx
time=“2021-02-03T12:02:08Z” level=debug msg=“Starting integrations group.” component=integrations.Manager file=/etc/newrelic-infra/integrations.d/jvm-metrics.yml
time=“2021-02-03T12:02:08Z” level=debug msg=“Found integration config file.” component=PluginRegistry configFile=/etc/newrelic-infra/integrations.d/jvm-metrics.yml
time=“2021-02-03T12:02:08Z” level=debug msg=“Ignoring v4 integration. To be loaded later.” component=PluginRegistry configFile=/etc/newrelic-infra/integrations.d/jvm-metrics.yml integration=
time=“2021-02-03T12:02:08Z” level=debug msg=“Running task.” component=integrations.Definition integration_name=nri-jmx
time=“2021-02-03T12:02:08Z” level=debug msg=“Running through all discovery matches.” component=integrations.Definition integration_name=nri-jmx
time=“2021-02-03T12:02:08Z” level=debug msg=“Waiting while the integration instances run.” component=integrations.runner.Runner env=dev integration_name=nri-jmx
time=“2021-02-03T12:02:08Z” level=debug msg=“Integration instances finished their execution. Waiting until next interval.” component=integrations.runner.Runner env=dev integration_name=nri-jmx

There’s a known issue affecting JMX integrations on the kubernetes integration that was fixed in the last release.
I’d recommend upgrading and seeing if that resolves the issue.

nri-kubernetes 2.2.0 release notes for reference.