Monitoring Frequency-Gateway with AWS CloudWatch

This guide explains how to set up monitoring for the Frequency-Gateway application using AWS CloudWatch for logging and metrics collection. CloudWatch offers in-depth metrics for system performance, container health, and more through Container Insights. For further details on CloudWatch setup, see the AWS CloudWatch Agent on Kubernetes documentation.

Prerequisites

  1. CloudWatch Agent: Install the CloudWatch Agent in your Kubernetes cluster, typically as a DaemonSet, to ensure metrics are collected from each node.
  2. IAM Roles: Ensure your cluster has the required permissions to write metrics and logs to AWS CloudWatch. Use IAM roles for service accounts or AWS IAM roles to attach permissions.

Step 1: Define CloudWatch Configuration in values.yaml

Customize your Helm chart's values.yaml file to enable CloudWatch and specify required parameters. Example:

cloudwatch:
  enabled: true
  region: "us-east-1"
  cluster_name: "MyCluster"
  collection_interval: 60
  enhanced_container_insights: true
  flush_interval: 5

Step 2: Create the CloudWatch ConfigMap

The ConfigMap defines the JSON configuration for the CloudWatch agent. The following Helm template dynamically generates the ConfigMap based on values in values.yaml:

{{- if .Values.cloudwatch.enabled }}
apiVersion: v1
kind: ConfigMap
metadata:
  name: cwagent-config
  namespace: amazon-cloudwatch
data:
  cwagentconfig.json: |
    {
      "agent": {
        "region": "{{ .Values.cloudwatch.region }}"
      },
      "logs": {
        "metrics_collected": {
          "kubernetes": {
            "cluster_name": "{{ .Values.cloudwatch.cluster_name }}",
            "metrics_collection_interval": {{ .Values.cloudwatch.collection_interval }},
            "enhanced_container_insights": {{ .Values.cloudwatch.enhanced_container_insights }}
          }
        },
        "force_flush_interval": {{ .Values.cloudwatch.flush_interval }}
      }
    }
{{- end }}

This configuration focuses on Kubernetes cluster-level insights, collecting metrics on an interval and enabling enhanced container insights.

Step 3: Deploy the ConfigMap

  1. Apply the ConfigMap with the configuration to the amazon-cloudwatch namespace in your Kubernetes cluster.
  2. Restart the CloudWatch agent pods to load the updated configuration.

Step 4: View Logs and Metrics in CloudWatch

After deploying the CloudWatch agent with the ConfigMap, access CloudWatch to view real-time logs and metrics under your designated log group and cluster name. For more specific metrics and alerts, refer to AWS documentation for configuring CloudWatch Alarms and Dashboards.

For further configuration details, refer to the AWS documentation for the CloudWatch Agent on Kubernetes.