lambda to eks pod migration

Hello, based on your description, I think KEDA may be helpful to you.

Here are some links I found that describe how to use KEDA to integrate SQS and scale it in EKS Pod:

• https://aws.amazon.com/cn/blogs/containers/scalable-and-cost-effective-event-driven-workloads-with-keda-and-karpenter-on-amazon-eks/
• https://github.com/aws-samples/event-driven-autoscaling-using-podidentity-and-keda/tree/main
• https://github.com/aws-samples/amazon-eks-scaling-with-keda-and-karpenter
• https://docs.aws.amazon.com/prescriptive-guidance/latest/patterns/event-driven-auto-scaling-with-eks-pod-identity-and-keda.html

Converting AWS Lambda functions to EKS pods requires careful consideration of several aspects:

1. For trigger points like EventBridge and SQS:

• You can implement controllers or operators in Kubernetes that listen to AWS events
• Use AWS Event Source for Kubernetes to connect event sources to your pods
• Implement a sidecar pattern where a container in your pod connects to AWS services and forwards events to your application container

2. For SQS polling and timeouts:

• Implement a consumer application in your pod that polls SQS queues
• Configure appropriate visibility timeouts in your application code
• Use backoff strategies for polling to avoid excessive API calls
• Implement proper error handling and dead-letter queues

3. For scaling:

• Use Horizontal Pod Autoscaler (HPA) to scale based on CPU/memory metrics
• Consider implementing custom metrics adapters to scale based on queue length
• Use Vertical Pod Autoscaler to optimize CPU and memory allocation
• For event-driven workloads, KEDA (Kubernetes Event-driven Autoscaling) can help scale based on event sources

When migrating from Lambda to EKS, consider these best practices:

• Start with a hybrid approach, gradually shifting traffic from Lambda to EKS
• Use TargetGroupBinding from AWS Load Balancer Controller if you need to maintain existing endpoints
• Containerize your Lambda code, ensuring it's properly adapted for continuous running rather than function invocation
• Consider using AWS Fargate for serverless container execution if you want to maintain a serverless operational model
• Implement proper health checks and readiness probes for your pods
• Set up appropriate monitoring and logging for your containerized applications

For controlled migration, you can:

• Deploy your containerized application on EKS
• Test thoroughly before shifting production traffic
• Implement a gradual traffic shift strategy
• Maintain the ability to roll back if issues arise
  Sources
  Effortless application modernization: migrate to Amazon EKS with existing NLB setup | Containers
  Get started with AWS Fargate for your cluster - Amazon EKS
  Place Kubernetes Pods on Amazon EKS by using node affinity, taints, and tolerations - AWS Prescriptive Guidance