BalancerLens

BalancerLens is a web app (with optional lightweight agents) that pinpoints why load balancers distribute traffic unevenly across targets. Instead of just showing high-level metrics, it correlates per-request attributes (headers, cookies, source ASN/region, TLS params, HTTP/2 vs HTTP/1.1, connection reuse) with backend selection outcomes to explain real-world imbalance: sticky sessions gone wrong, NAT-induced source IP collapse, uneven connection pooling, misweighted target groups, or health-check flapping. It ingests logs and metrics from common stacks (AWS ALB/NLB, NGINX, Envoy, HAProxy, Kubernetes Ingress) and produces a “skew report” with ranked root-cause hypotheses and concrete fixes. This is not a silver bullet: it won’t replace your load balancer or magically optimize latency. It’s a diagnostic tool that saves hours of war-room time when traffic distribution looks ‘fine’ in dashboards but is burning specific nodes or regions.