$ k8scalc compare traefik haproxy ingress
Traefik vs HAProxy Ingress
Traefik vs HAProxy Kubernetes ingress controllers — configuration UX, performance, TCP/UDP routing, middleware, and use case fit for self-hosted clusters.
FeatureTraefikHAProxy Ingress
Configuration model
IngressRoute CRDs
ConfigMap + annotations
Auto service discovery
Yes (label-based)
No
Built-in dashboard
Yes—
Yes (stats page)—
Let's Encrypt built-in
Yes (native)
No (via cert-manager)
Middleware / plugins
Middleware CRD chain
Limited
Raw TCP throughput
High (Go)
Highest
Load balancing algorithms
Round-robin, WRR
Round-robin, leastconn, etc.
Resource usage (idle)
~30 MB RAM
~20 MB RAM
TCP/UDP routing
Yes (IngressRouteTCP/UDP)—
Yes (native)—
Community & ecosystem
Large
Smaller
gRPC support
Yes (native)—
Yes—
Verdict
Traefik wins on developer UX — automatic discovery, IngressRoute CRDs, built-in Let's Encrypt, and a dashboard make it the easiest ingress to operate. HAProxy wins on raw performance and advanced load balancing algorithms. For self-hosted clusters where ops simplicity matters, Traefik. For high-throughput, latency-sensitive, or financial workloads, HAProxy.