Optimize AWS Proxy Deployment

I already have a functioning codebase that reaches Amazon endpoints, but I now need to harden and streamline the network layer. The top priority is Optimized Proxy Deployment: spinning up several proxy servers on AWS, each tied to its own Elastic IP and placed in regions nearest to the Amazon infrastructure we query. The target footprint is roughly 100 static IPs—half reserved for one domain and half kept in a common pool. Every outbound call must rotate to the next address so no two consecutive requests share the same IP, keeping average round-trip latency between 150 ms and 300 ms. When Amazon responds with a 429, the system should swap both the IP and candidateId instantly, with zero observable delay to the application. A second goal is lowering the overhead around our create-application / WebSocket handshake; I’m happy to hear your ideas on how feasible that reduction is inside AWS or via third-party tooling. Deliverables • Infrastructure code (Terraform, CloudFormation or similar) to provision and recycle the proxy fleet, Elastic IPs and security groups • Updates to my existing repository to hook into the new rotation and failover logic • Documentation covering deployment, scaling, cost considerations and how to expand beyond 100 IPs Acceptance criteria 1. Round-robin IP rotation on every single request 2. Mean latency consistently stays within 150-300 ms under agreed test load 3. Automatic failover resolves a 429 in under one second 4. Infrastructure spins up and tears down cleanly using only the supplied scripts Please highlight your relevant experience building high-throughput proxy or scraping architectures on AWS, especially if you have worked with Squid, HAProxy, custom reverse proxies, Elastic IP management, auto-scaling and multi-region optimizations.

Project ID: 40274043