Implementation of Nginx load balancing cluster

(1) Experimental environment

youxi1 192.168.5.101 Load Balancer

youxi2 192.168.5.102 Host 1

youxi3 192.168.5.103 Host 2

(2) Nginx load balancing strategy

Nginx load balancing is used to select a server from the backend server list defined in the upstream template to receive user requests. A basic upstream module is as follows:

upstream [server group name]{
  server [IP address]:[port number];
  server [IP address]:[port number];
  ....
}

After the upstream module is configured, you need to specify the access reverse proxy to the server list in the following format:

location ~ .*$ {
  index index.jsp index.html;
  proxy_pass http://[server group name];
}

Extension: nginx location configuration rules: https://www.jb51.net/article/182472.htm

This completes the most basic load balancing, but it cannot meet actual needs. Currently, Nginx's upstream module supports six load balancing strategies (algorithms): polling (default method), weight (weight method), ip_hash (based on IP allocation method), least_conn (least connection method), fair (response time method provided by a third party), and url_hash (based on URL allocation method passed by a third party).

1) Polling

The most basic configuration method is the default load balancing strategy of the upstream module. Each request will be evenly distributed to different backend servers in chronological order. The parameters are as follows:

Note: 1. Servers marked as down will be automatically removed; 2. Polling is the default; 3. This strategy is suitable for services with stateless server configuration and short flat blocks

2) weight

Weighted method, specifies the probability of polling based on the polling strategy. It can also be considered that a weight parameter is added on the basis of polling. This parameter specifies the probability of polling and its value is number. The upstream module configuration template is as follows:

upstream [server group name]{
  server [IP address]:[port number] weight=2;
  server [IP address]:[port number];
  ....
}

In this example, the default value for a server without a weight parameter is 1. The weight value is proportional to the access ratio. The sum of all weight values ​​is a cycle unit, and the server's own weight value is the number of polls within the cycle unit.

Note: 1. The higher the weight, the more requests are allocated; 2. This strategy can be used in combination with the least_conn strategy and the iphash strategy; 3. This strategy is more suitable for situations where there is a large difference in server hardware configuration.

3) ip_hash

According to the IP allocation method, the load balancer is specified to be allocated based on the client IP. This method ensures that the same client request is consistently sent to the same server to ensure the session session. In this way, each visitor accesses a fixed backend server, which can solve the problem that the session cannot cross servers. The upstream module configuration template is as follows:

upstream [server group name]{
  ip_hash;
  server [IP address]:[port number] weight=2;
  server [IP address]:[port number];
  ....
}

Note: 1. Versions prior to nginx 1.3.1 cannot use weights in ip_hash; 2. ip_hash cannot be used with backup at the same time; 3. This strategy is suitable for stateful service programs, such as session; 4. When a server needs to be removed, it must be manually shut down.

4) least_conn

The least connection mode sends the request to the backend server with the least number of connections. Polling distributes requests evenly to each backend so that their loads are roughly the same. However, some requests take a long time and cause a high load on the backend where they are located. In this case, least_conn can achieve better load balancing effect. The upstream module configuration template is as follows:

upstream [server group name]{
  least_conn;
  server [IP address]:[port number] weight=2;
  server [IP address]:[port number];
  ....
}

Note: This strategy is suitable for server overload situations where requests with varying processing times take varying amounts of time.

5) fair

Response time method: requests are allocated according to the response time of the server, with requests with shorter response time being given priority. The upstream module configuration template is as follows:

upstream [server group name]{
  server [IP address]:[port number] weight=2;
  server [IP address]:[port number];
  ....
  fair;
}

Note: A third-party plugin needs to be installed.

6) url_hash

The URL distribution method distributes requests according to the hash result of the accessed URL, so that each URL is directed to the same backend server. It should be used in conjunction with cache hits. Multiple requests for the same resource may arrive at different servers, resulting in unnecessary multiple downloads, low cache hit rates, and a waste of some resource time. By using url_hash, the same URL (that is, the same resource request) can reach the same server. Once the resource is cached, it can be read from the cache when the request is received again. The upstream module configuration template is as follows:

upstream [server group name]{
  hash $request_uri;
  server [IP address]:[port number] weight=2;
  server [IP address]:[port number];
  ....
}

Note: 1. A third-party plug-in needs to be installed; 2. uri is i, not lowercase L.

(3) Experiment

1) Compile and install nginx on the load balancer youxi1

Install nginx dependencies

[root@youxi1 ~]# yum -y install gcc gcc-c++ autoconf automake zlib zlib-devel openssl openssl-devel pcre pcre-devel

Upload the nginx source package nginx-1.14.1.tar.gz, decompress and install

[root@youxi1 ~]# tar xf nginx-1.14.1.tar.gz -C /usr/local/src/
[root@youxi1 ~]# cd /usr/local/src/nginx-1.14.1/
[root@youxi1 nginx-1.14.1]# ./configure --prefix=/usr/local/nginx --with-http_dav_module --with-http_stub_status_module --with-http_addition_module --with-http_sub_module --with-http_flv_module --with-http_mp4_module
[root@youxi1 nginx-1.14.1]# make && make install
[root@youxi1 nginx-1.14.1]# echo $?
0

Parameter Description:

--with-http_dav_module, enable ngx_http_dav_module support (add PUT, DELETE, MKCOL: create collection, COPY and MOVE methods) by default, it is disabled and needs to be compiled in;

--with-http_stub_status_module, enable ngx_http_stub_status_module support (get nginx working status since the last startup);

--with-http_addition_module, enable ngx_http_addition_module support (as an output filter, supports incomplete buffering and partial response requests);

--with-http_sub_module, enable ngx_http_sub_module support (allows to replace some text in nginx response with some other text);

--with-http_flv_module, enable ngx_http_flv_module support (provides seeking to memory using time-based offset files);

--with-http_mp4_module, enable support for mp4 files (provides seeking to files using time-based offsets in memory).

Generate nginx user

[root@youxi1 nginx-1.14.1]# useradd -M -s /sbin/nologin nginx

Start and add auto-startup

[root@youxi1 nginx-1.14.1]# /usr/local/nginx/sbin/nginx
[root@youxi1 nginx-1.14.1]# echo /usr/local/nginx/sbin/nginx >> /etc/rc.local
[root@youxi1 nginx-1.14.1]# chmod +x /etc/rc.d/rc.local

If the firewall is turned on, remember to add the port number

[root@youxi1 nginx-1.14.1]# firewall-cmd --permanent --zone=public --add-port=80/tcp && firewall-cmd --reload
success
success

Check the interface to see if nginx is installed normally

After the test is completed, modify the nginx configuration file and finally restart nginx

[root@youxi1 nginx-1.14.1]# cp /usr/local/nginx/conf/nginx.conf{,.bak}
[root@youxi1 nginx-1.14.1]# vim /usr/local/nginx/conf/nginx.conf
user nginx; //line 2 location / { //line 43 onwards root html;
  index index.html index.htm;
  if ($request_uri ~* \.html$){
    proxy_pass http://htmlservers;
  }
  if ($request_uri ~* \.php$){
    proxy_pass http://phpservers;
  }
  proxy_pass http://picservers;
}
upstream htmlservers { //Under the http module, add server 192.168.5.102:80 at the same level as the server module;
  server 192.168.5.103:80;
}
upstream phpservers{
  server 192.168.5.102:80;
  server 192.168.5.103:80;
}
upstream picservers {
  server 192.168.5.102:80;
  server 192.168.5.103:80;
}
[root@youxi1 nginx-1.14.1]# /usr/local/nginx/sbin/nginx -s reload

2) Deploy the web page program on youxi2 and youxi3

[root@youxi2 ~]# yum -y install httpd
[root@youxi2 ~]# echo youxi2 > /var/www/html/index.html
[root@youxi2 ~]# echo youxi2.php > /var/www/html/index.php
[root@youxi2 ~]# echo youxi2.other > /var/www/html/index.jsp
[root@youxi2 ~]# systemctl start httpd.service
 
[root@youxi3 ~]# yum -y install httpd
[root@youxi3 ~]# echo youxi3 > /var/www/html/index.html
[root@youxi3 ~]# echo youxi3.php > /var/www/html/index.php
[root@youxi3 ~]# echo youxi3.other > /var/www/html/index.jsp
[root@youxi3 ~]# systemctl start httpd.service

If the firewall is turned on, remember to add the port number

[root@youxi2 ~]# firewall-cmd --permanent --zone=public --add-port=80/tcp && firewall-cmd --reload
success
success
 
[root@youxi3 ~]# firewall-cmd --permanent --zone=public --add-port=80/tcp && firewall-cmd --reload
success
success

3) Testing

First test the html page

Then test the PHP page

Finally test other pages

Reference: https://www.cnblogs.com/1214804270hacker/p/9325150.html

This is the end of this article about how to build a load balancing cluster with Nginx. For more information about Nginx load balancing cluster, please search for previous articles on 123WORDPRESS.COM or continue to browse the following related articles. I hope you will support 123WORDPRESS.COM in the future!