Feature Request: Add Service Mesh Support to AWS EKS Integration

[random-letter-generator]

Problem

Currently the AWS integration does not support service mesh, thus lacking the tracking of API request latency data.Currently, the AWS integration in BottleNetes does not support service mesh (e.g., Istio or others). This limitation prevents the tool from tracking API request latency data for pods running in AWS EKS clusters. While the local Kubernetes setup supports service mesh, allowing for comprehensive observability of pod latencies, the AWS integration requires additional setup, resulting in a lack of feature parity.

Solution

Adding support of service mesh (Istio or others) to the AWS integration.Add service mesh support (e.g., Istio or similar frameworks) to the AWS EKS integration within BottleNetes.

• Use Cases:

  • Enable BottleNetes to monitor API request latency data for individual pods running in AWS EKS clusters.
  • Provide consistent observability features across local Kubernetes clusters and AWS EKS clusters.
  • Simplify user setup by integrating service mesh configuration into the AWS integration flow.

• Implementation Suggestions:

  • Automatically detect and configure service mesh for EKS clusters when connecting via the AWS CLI and AWS EKS CLI tools.
  • Provide users with documentation or an optional setup script to enable Istio or compatible service mesh frameworks on their AWS EKS clusters.
  • Fetch and display latency metrics for each pod in the BottleNetes dashboard UI, similar to the current local Kubernetes integration.

• Alternative Solutions Considered:

  • Requiring users to manually set up and configure a service mesh on their AWS EKS clusters. However, this approach is less user-friendly and may deter users from utilizing this functionality.

Additional information

Current Limitation:
When BottleNetes is hooked to an AWS EKS cluster, it cannot fetch API request latency data for individual pods due to the absence of service mesh support.

👨‍👧‍👦 Contributing

• 🙋‍♂️ Yes, I'd love to make a PR to implement this feature!