Load Balancing Across Backends with NGINX upstream
Use NGINX upstream and proxy_pass to distribute traffic across multiple backend servers, with passive health checks and header forwarding.
ReadMeFeed / Community knowledge
Real questions. Useful conversations. Find the people who know your stack.
Use NGINX upstream and proxy_pass to distribute traffic across multiple backend servers, with passive health checks and header forwarding.
NGINX’s proxy_cache turns the reverse proxy into a disk‑backed cache. Learn how to define the cache zone, control TTLs, verify hits, and manage disk space and purging in this practical guide.
Configure NGINX rate limiting with ngx_http_limit_req_module: define shared memory zones, apply burst and nodelay, handle proxied client IPs, and avoid common placement mistakes.
Learn how NGINX's limit_req module uses a leaky‑bucket filter to throttle API traffic, with a concrete configuration, trade‑offs, and verification steps.
Learn how to diagnose and fix Socket.io connection failures where clients get stuck in 'polling' mode due to CORS, proxy header issues, or missing sticky sessions.
Learn how to deploy and manage isolated web applications using Yunohost's automated reverse proxy and system-user isolation to prevent lateral movement.
Learn how to safely install newer software on Debian stable by enabling the Backports repository, with a step‑by‑step nginx upgrade example and practical tips for maintenance.
How to configure NGINX limit_req with a shared-memory zone, burst queue, and real-time monitoring via stub_status to tame traffic spikes without silently dropping requests.
NGINX's limit_req module smooths abusive traffic with a leaky bucket — but the defaults (503 rejections, no burst, IP keys behind a load balancer) are subtly wrong for most APIs. Here's a config that works and how to verify it.