TL;DR - Crunch numbers on peak traffic to understand overall HTTP content length metrics. Nginx is commonly used as an ingress proxy - your public load balancer receives a request and passes it on to Nginx, which in turn checks it against pre-set routing rules and passes it on to the right application. In my setup, the traffic pattern consisted of HTTP requests but I had to enable my public endpoint temporarily to receive Prometheus remote write metric traffic. As soon as I did that, I saw a flurry of these warning log lines in Nginx:
Calculate HTTP content-length metrics on cli
Calculate HTTP content-length metrics on cli
Calculate HTTP content-length metrics on cli
TL;DR - Crunch numbers on peak traffic to understand overall HTTP content length metrics. Nginx is commonly used as an ingress proxy - your public load balancer receives a request and passes it on to Nginx, which in turn checks it against pre-set routing rules and passes it on to the right application. In my setup, the traffic pattern consisted of HTTP requests but I had to enable my public endpoint temporarily to receive Prometheus remote write metric traffic. As soon as I did that, I saw a flurry of these warning log lines in Nginx: