Cherry-pick of PR #25640 to 7.x branch. Original message:

What does this PR do?

This PR changes how kubernetes module handle state_* metricsets which share same target endpoint. The idea originates from https://github.com/elastic/beats/blob/master/x-pack/metricbeat/module/cloudfoundry/cloudfoundry.go

Note: At this point the PR stands more like a PoC with changes only applied at state_container and state_pod metricsets.

If we agree with this solution (and make sure that it would be applied with Agent too) we can extend it to the rest of state_* metricsets as well as to metricsets fetch from kubelet's endpoint (node, pod, container, volume, system)

In upcoming PR we will apply similar strategy for metricsets fetching from kubelet's endpoint (node, pod, container, volume, system)

Why is it important?

To improve the performance of the module by avoid fetching same content multiple times.

Checklist

• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have made corresponding change to the default configuration files
• I have added tests that prove my fix is effective or that my feature works
• I have added an entry in CHANGELOG.next.asciidoc or CHANGELOG-developer.next.asciidoc.

Author's Checklist

• Does it have impact for Agent deployments too?
  Verified with different module's config blocks. See step 4 of testing notes below.

How to test this PR locally

1. On a k8s cluster install kube_state_metrics (https://github.com/kubernetes/kube-state-metrics/tree/master/examples/standard)
2. Expose kube_state_metrics endpoint on the local machine -> kubectl -n kube-system port-forward svc/kube-state-metrics 8081:8080
3. Start capturing traffic on kube-state-metrics' endpoint -> tcpdump -i any -s 0 'tcp[((tcp[12:1] & 0xf0) >> 2):4] = 0x47455420'
4. Test with different module's config blocks (what Agent does):

- module: kubernetes
  metricsets:
    - state_pod
  period: 10s
  hosts: ["0.0.0.0:8081"]
  add_metadata: false

- module: kubernetes
metricsets:
- state_container
period: 10s
hosts: ["0.0.0.0:8081"]
add_metadata: false

Verify with tcpdump's output that only one request takes place no matter how many modules/metricsets are enabled.

5. Test one combined module block:

- module: kubernetes
  metricsets:
    - state_pod
    - state_container
  period: 10s
  hosts: ["0.0.0.0:8081"]
  add_metadata: false

Verify with tcpdump's output that only one request takes place no matter how many modules/metricsets are enabled.

Related issues

• Relates Refactor kubernetes metricsets to use single calls per endpoint #24869