From 6db7f62c1df3d11a4fccd3e654b2d425a0c1fa10 Mon Sep 17 00:00:00 2001 From: cn Date: Sat, 10 Nov 2018 10:22:46 +0100 Subject: [PATCH] Add more details to README --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index e21e15c..6777766 100644 --- a/README.md +++ b/README.md @@ -4,7 +4,7 @@ This is a Lua plugin for the Nginx web server that automatically collects and su In constrast to commercial and proprietary solutions such as [Luameter](https://luameter.com/) or [NGINX Plus](https://www.nginx.com/products/) with it's [ngx_http_status_module](http://nginx.org/en/docs/http/ngx_http_status_module.html), this plugin is open source software while featuring more and additional metrics compared to those available via the open source [ngx_http_stub_status_module](http://nginx.org/en/docs/http/ngx_http_stub_status_module.html). (Yea, [other web servers](https://redmine.lighttpd.net/projects/1/wiki/Docs_ModStatus) deliver more information by default...) -This plugin takes inspiration from other Nginx metric libraries like [nginx-lua-prometheus](https://github.com/knyar/nginx-lua-prometheus) but differs fundamentally in the metrics submission handling by. In certain intervals it automatically pushes metrics to the configured Graphite (i.e. Carbon) host using pure Lua code instead of exposing them via a separate web page for polling. +This plugin takes inspiration from other Nginx metric libraries like [nginx-lua-prometheus](https://github.com/knyar/nginx-lua-prometheus) but differs fundamentally in the metrics submission handling. In certain intervals it automatically pushes metrics to the configured instance(s) of [Carbon](https://github.com/graphite-project/carbon/) using pure Lua code instead of exposing them via a separate web page for HTTP polling. For that is uses the Graphite plaintext protocol over TCP establishing a new connection for every push. The metrics collection happens on every request for which the user configures a suitable `log_by_lua` direcitve and towards server-wide global counters (finer granularity might be added later). The counters are realized using a single shared dictionary across all Nginx worker threads which has constant memory usage (128 KiB currently, may be reduced further).