nginx is an HTTP, reverse proxy server & TCP/UDP proxy. written by Igor Sysoev using C language. nginx help many web apps to made them fast like yandex, VK, netflix ..etc.
To install nginx it depend on the OS you are runing, to do so you can use packages on linux or download the zip file from downloads for windows & linus.
To install nginx on MacOS will use brew & runing this command on cli
brew install nginxOn Ubuntu OS it must run some command to install nginx
First, will install required packages:
sudo apt install curl gnupg2 ca-certificates lsb-releaseSecond, config apt repo for stable nginx packages, will run:
echo "deb http://nginx.org/packages/ubuntu `lsb_release -cs` nginx" \| sudo tee /etc/apt/sources.list.d/nginx.list
Third, will import an official signing key to allow apt to verify the packages authenticity:
curl -fsSL https://nginx.org/keys/nginx_signing.key | sudo apt-key add -You can verify that with:
sudo apt-key fingerprint ABF5BD827BD9BF62the output can look like this:
pub rsa2048 2011-08-19 [SC] [expires: 2024-06-14]
573B FD6B 3D8F BC64 1079 A6AB ABF5 BD82 7BD9 BF62
uid [ unknown] nginx signing key <signing-key@nginx.com>Forth & finaly, will runing the following command to install it:
sudo apt update && sudo apt install nginxThat it for ubuntu & now you can have a fun with configuring it.
Note:
For more detail on how to install nginx on other Linux destribution OS you can visit this link for more detail. To verify if nginx install run
nginx -vAfter installing nginx, now will build a web server using nginx to serve a content. To do that will edit the default config file.
Nginx has 1 master proccess & several worker processes. The main process read & evalute configuration & maintain worker processes & in turn workers do the processing of requests. Nginx embrasse event-driven model & OS-dependent mechanisms to efficiently distribute requests among workers. The number of workers can defined in the configuration & may be fixed/automatically adjusted to the number of available CPU cores for more visit this link.
To lunch nginx run the executable file or run this command:
nginxand to stop nginx from runing it can be done in 2 ways: the first one is stoping nginx after finishing the processes of all current requests via
nginx -s quitthe second one is to shutdown all worker processes using:
nginx -s stopNote:
If you notice the -s param are used on both that mean passing a signal to the master process
to execute an action passed after -s it can be one of stop, quit, reopen, reload
will use the 2 others in next sections.
To list all nginx processes runing you can use
ps -ax | grep nginxNginx use a simple language to edit it configuration based on directives. There is 2 kind of directives,
simple directive which consider as an instruction look like name_of_directive param;, end with ;,
the socond one is block directive(called context also) which is similar to simple directive instead of ;
it end with { & contain other directives(instructions).
The way nginx & its modules work is determined in the configuration file. By default, the configuration
file is named nginx.conf & placed in the directory /usr/local/nginx/conf, /etc/nginx,
or /usr/local/etc/nginx.
To create a static content server will edit the nginx.conf file located in one of
/usr/local/nginx/conf, /etc/nginx, or /usr/local/etc/nginx(for linux & MacOS users), but before
that will create 2 folders /data/www & /data/images to put html files & images respectively
for example put create a file index.html contain:
<html>
<head>
<title>Serving static content with Nginx</title>
</head>
<body>
<h1>This file served with NGINX</h1>
</body>
</html>In /data/images/ put some images like example.png to serve it later using nginx, then Open the
configuration file & replace its content with:
http {
server {
listen 8080;
location / {
root /data/www;
}
location /images {
root /data;
}
}
}
events {} # ignore this directive but it requiredNow you can re/start nginx with
# in case nginx was shut
nginx
# or see in next section for more detail to reload configuration
nginx -s reloadvisit http://localhost:8080/ for html page or http://localhost:8080/images/example.png for the image. So let explain what all that mean:
httpcontext(block directive contain other blocks) to servehttprequestsservercontext to specify the configuration of virtual serverlisteninform nginx the number of port to serve contentlocationsets configuration depending on a request URI, in our case there is 2 location blocks the first one tell nginx to serve all request prefixed with/& the secondth to serve all request containing/imagesin url for example serve all request ofhttp://localhost:8080/images/example.pngrootthe path where to serve the request.eventsrequired by nginx let ignore it for now.
Note:
For matching requests, the URI will be added to the path specified in the root directive, that is, to /data/www.
If there are several matching location blocks nginx selects the one with the longest prefix.
After editing the configuration file on /usr/local/etc/nginx/nginx.conf you can stop & start
nginx in 2steps or simply reload it using:
nginx -s reloadWhen reload the configuration nginx verify if the new configuration are valid, if it is then the master process start new worker processes & send a signal to old ones to shutdown after serving all current request & not accepting new connections. Otherwise the master process roll back & keep using the old configuration. For more control of nginx.
To create a proxy server with nginx it realy simple by creating new directory /data/proxy-server
containing html files then edit the config file & add new server context to the previous configuration like:
server {
listen 8888;
root /data/proxy-server;
location / {}
}that mean, serve a request from port 8888 & files from /data/proxy-server then edit the server
contexts from previous configuration to be:
server {
listen 8080;
location {
proxy_pass http://localhost:8888;
}
location ~ \.(gif|jpg|png)$ {
root /data/images;
}
}if you notice we add new directive proxy_pass, this send an instruction to nginx to inform it
to proxy all requests from http://localhost:8080 to the new url http://localhost:8000,
in other words if you vist http://localhost:8080 nginx serve you http://localhost:8000,
& second location directive is start with ~ which mean is a regular expression to math one
of those extensions gif or jpg or png & serve them from /data/images so if you visit
http://localhost:8080/example.png will work but not http://localhost:8000/example.png;
Let assume that your config file grows & annoyes you to handle it in one file so you can use separate
your configuration to several files & include it using include directive, which is of type
simple(not block directive), to put your configuration files following this syntaxe:
# here you can write some configuration on main context
# here include external config which contain http directive
include http.conf
# events directive
events {} Another directive which can be in main context is worker_processes which accept as a
parametre a number or auto to specify the number of worker processes to use, however those
number can related to some condition like a number of CPU core, the number of hard disk
drives that store data, & load pattern.
In some cases, you want to debug your server or see errors nginx fired. Nginx give you
a possibility to specify a custom path to file who include those errors/debug logs using
error_log, which accept 2 parametre the first one is the path to given file, if it dosn't
exist yet nginx will created for you(not the folder just files) & the second one is the type
of logs to save in that file it one of debug, info, notice, warn, error, crit, alert, or
emerg, if you specified error the log file will contain error, crit, alert, or emerg logs too.
This directive can be in thos contexts main, http, mail, stream, server, location.
Caching is a weapon with 2 side, it can be good & bad if you dont know how to exploite it. Caching
can enhance your application performance & take it to the next level, so in this section will use
nginx to cache some content. To do that its very simple using 3 directives. First, let create a proxy
server with:
http {
server {
listen 8888;
location / {
root /Users/mac/Desktop/nginx/proxy-server;
}
}
server {
listen 8080;
location / {
proxy_pass http://localhost:8888;
}
}
}
events {}for now there is nothin new. Second, let user proxy_cache_path to define the path where save data, size of cache, max size of cache so add the following configuration in http context before 2nd server context:
#...
proxy_cache_path /data/cache levels=1:2 max_size=1g keys_zone=app_cache_zone_name:10m use_temp_path=off inactive=60m:
# ...this inform nginx to save cache in /data/cache or create it if not exist with 2 levels of subfolders
(best practice is to kepp this on 1:2 or less) with the maximum size of 1 Giga, if it exceed nginx will
remove the least used cached data, then save the cache meta data on the zone app_cache_zone_name with
size of 10 Miga. The content cached with nginx created imediatly without passing with temporare folder
when using use_temp_path=off param & refresh those cached content if not reached in 60 minutes. Howeer
this cache not enabled yet, to do so will add those directive in 2nd server context:
server {
listen 8080;
location / {
proxy_cache app_cache_zone_name;
proxy_cache_valid 200 10m;
proxy_pass http://localhost:8888;
}
}those new directives mean to enable the caching zone app_cache_zone_name used as param of proxy_cache
& proxy_cache_valid to inform nginx to cache all http requests content served with status of 200 for 10 minutes. Finally
you can reload the configuration by sending a signal of HUP using:
nginx -s reload& just like that you implemented a caching using nginx to emprove your app performance.
Some times errors happen or server goes down & your server are not accessible then you app users(customers) see an ugly error message informing them the app are down & not reachable in this moment or maybe your app got a spike trafic, to avoid this situation you can use cache to serve a content for users using proxy_cache_use_state where your enabling you cache like:
# ...
location / {
proxy_cache app_cache_zone_name;
proxy_cache_valid 200 10m;
proxy_cache_use_stale error timeout http_500 http_501 http_504;
proxy_pass http://localhost:8888;
}
# ...this directive inform nginx to serve a cached content in case of error ,by adding error, param in proxied
servers or when all worker processes are busy serving other user via timeout or when proxied servers
responsd with status 500, 501 & 504.
There is many directives you can use to enhance & customize nginx to cache
requests & content of you web application. In this section will illustrate
some of those directives might helps you
proxy_cache_keysdefines a key for caching, for exampleproxy_cache_key "$host$request_uri $cookie_user";becausenginxuse the uri of request(hashed using md5) to cache the requests & this directive informnginxthe cache name of each request so be carefull when using it, it can be tricky if you missed & cache 2 different requests with in one directory & cache name.proxy_cache_min_usesaccept as param anumberwhich is a number of requests received to cache it as an exampleproxy_cache_min_uses 3instructsnginxto cache content when it receive 3 requests to that uri.proxy_cache_methodsto specify whichhttprequest method will cache, so it value will be an http method.proxy_cache_bypasseused to bypasse cache referred to as “punching a hole” & get content from origin server rather than cache
Note:
You can send a custom header to inform if the response content served from cache using add_header
directive & $upstream_cache_status global variable at location context as following:
add_header X-Cache-Status $upstream_cache_status;the response will contain X-Cache-Status header with one of MISS, BYPASS, EXPIRED, STALE or
HIT where means respectivelly not served from cache, passed(no cache for this request), cache are expired,
serve cache version where the proxied server are down/timeout & HIT server from cache. For more details & FAQ you can visit this source.