How to build a load balancing server based on Centos7+Nginx+Tomcat8

After manually installing nginx, everything is normal, and there are no problems with the configuration of the virtual host and load balancing. It was successfully enabled. The specific configuration file is as follows:

Added: include vhost/*.conf; to the nginx.conf file of conf to introduce all the .conf files in the vhost directory for future configuration. Simple, of course, you can add them one by one if necessary.

Because the load service is api service, the name of the conf file here is called api.xxx.com.conf. The configuration of opening the file is as follows:

The ip polling method is not used here, but the usstream method is used. For the specific differences between these methods, you can Baidu or Google for relevant introductions. It should be noted that proxy_pass is configured here as the name of upstream. Others are not allowed to be used. If the IP address is configured directly, load balancing will not be achieved.

The weight represented by weight, the larger the number, the higher the weight, the greater the chance of allocation. backup represents the hot standby server, that is, when there are problems with the first two loads, the hot standby server These visits will be automatically assumed.

The configuration of nginx load balancing configurator is completed at this point, but after the load is involved, if it is an ordinary URL, the session loss will occur when accessing, so at this time, it is necessary to deal with the session loss. Problem, to deal with this problem, there are multiple solutions. One is to use cookies proxy, and the other is to realize session sharing. Here, session sharing is used. When realizing sharing, relevant configurations need to be performed in tomcat. To achieve this According to requirements, the shared server of the session configured here is redis memory service, mainly to ensure efficient reading and storage of data. When using redis for session sharing, the following names must be added:

These packages support tomcat8 without any problems. No relevant tests have been conducted on the specific support of other versions.

Then modify the configuration file in tomcat:

The addresses of 10.0.0.x seen here are all completed within the internal network. The network is blocked, and only the server responsible for the load is directly connected to the external network.

At this time, the tomcat server and nginx server can be started successfully, but before that, the nginx server must be started first, otherwise tomcat may fail to start. The nginx service can be started smoothly

The three configurations are as follows:

The reference content added in nginx.conf

include vhost/*.conf;

Load balancing configuration:

upstream api.xxx.com {
 server 10.0.0.12:8080 weight=5;
 server 10.0.0.15:8081 weight=5 ;
 server 10.0.0.16:8080 weight=5 backup;
}
server {
 listen    80;
 server_name api.xxx.com;
 location /
  {
   proxy_set_header host $host;
   proxy_set_header x-real-ip $remote_addr;
   proxy_set_header x-forwarded-for $remote_addr;
   proxy_pass http://api.xxx.com;
    }
}

tomcat configuration

<valve classname="com.radiadesign.catalina.session.redissessionhandlervalve" />
<manager classname="com.radiadesign.catalina.session.redissessionmanager" 
       host="10.0.0.16" 
       port="10000" 
       database="0" 
       maxinactiveinterval="60" />

The above is the detailed content of How to build a load balancing server based on Centos7+Nginx+Tomcat8. For more information, please follow other related articles on the PHP Chinese website!