Improve cluster detection for components during translation

[musa-asad]

Description of the issue

When running on a Kubernetes cluster, the only way the CloudWatch Agent can detect the cluster name is:

1. If it's provided in the configuration:

"kubernetes": {
  "cluster_name": "<cluster-name>",
  "enhanced_container_insights": true
},
"application_signals": {
  "hosted_in": "<cluster-name>"
}

2. By making a ec2:DescribeTags call.

Ideally, the agent should minimize external API calls, so we should implement a global way to retrieve the cluster name in case it isn't provided in the configuration. We can fetch the K8S_CLUSTER_NAME environmental variable, which is populated by the Amazon CloudWatch Observability EKS add-on / Helm chart.

Description of changes

Important

Co-PR: aws-observability/helm-charts#167

Revision 1

• Create GetHostedIn function, which looks for cluster name under app signals.
• Create GetClusterName function, which looks for cluster name under kubernetes, in K8S_CLUSTER_NAME environmental variable, or by a ec2:DescribeTags call.
• Reference these functions in entity and app signals processors.
• Organize entity processor so that it only calls these functions when running in Kubernetes.
• Add unit tests for entity processor.

Revision 2

• Add GetClusterName() call for containerinsights.
• Add precedence testing.

License

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

Tests

In an EKS cluster, I added a print statement to verify the environmental variable's value is accessed during translation:

Requirements

Before commit the code, please do the following steps.

1. Run make fmt and make fmt-sh
2. Run make lint