Log segmentation
1. Too many log files in the enterprise server cause the following problems:
1. It is inconvenient for development and operation and maintenance to query too large log files, especially when overtime workers are tired;
2. Long time ago, log files were almost worthless, but manual cleaning was too cumbersome.
At this time, a solution is needed to automatically split the logs. The divided logs are not only "clean", but also convenient for regular log cleaning.
Experiment
2. Compile and install Nginx service
1. Obtain the source package on Windows remotely and mount it on Linux
[root@localhost ~]# smbclient -L //192.168.235.1 Enter SAMBA\root's password: Sharename Type Comment --------- ---- ------- LNMP Disk [root@localhost ~]# mkdir /abc [root@localhost ~]# mount.cifs //192.168.235.1/LNMP /abc Password for root@//192.168.235.1/LNMP: [root@localhost ~]# ls /abc Discuz_X3.4_SC_UTF8.zip nginx-1.12.0.tar.gz php-7.1.10.tar.bz2 mysql-boost-5.7.20.tar.gz nginx-1.12.2.tar.gz php-7.1.20.tar.gz
2. Decompress the package
[root@localhost ~]# cd /abc [root@localhost abc]# tar zxvf nginx-1.12.0.tar.gz -C /opt [root@localhost abc]# ls /opt nginx-1.12.0 rh
3. Install and compile component package
[root@localhost abc]# cd /opt [root@localhost opt]# yum install -y \ > gcc \ //C language > gcc-c++ \ //c++ language > pcre-devel \ //pcre language tools > zlib-devel //Compress function library
4. Create program users and configure related components of Nginx service
[root@localhost opt]# useradd -M -s /sbin/nologin nginx //Create program user nginx and restrict it to not log in terminal [root@localhost opt]# cd nginx-1.12.0/ [root@localhost nginx-1.12.0]# ./configure \ //Configure nginx > --prefix=//usr/local/nginx \ //Specify installation path > --user=nginx \ //Specify user name > --group=nginx \ //Specify the group to which the user belongs > --with-http_stub_status_module //Installation status statistics module
5. Compilation and installation
[root@localhost nginx-1.12.0]# make && make install
6. Optimize the startup script of Nginx service and establish the command soft link
[root@localhost nginx-1.12.0]# ln -s /usr/local/nginx/sbin/nginx /usr/local/sbin/ //Create nginx service command soft link to system command [root@localhost nginx-1.12.0]# systemctl stop firewalld.service //Turn off firewall [root@localhost nginx-1.12.0]# setenforce 0 //Turn off enhanced security [root@localhost nginx-1.12.0]# nginx //Enter nginx to start the service [root@localhost nginx-1.12.0]# netstat -ntap | grep 80 / / check port 80 of the service, and it is shown that it is enabled tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN 7520/nginx: master
7. System CTL management nginx script
[root@localhost ~]# vim /lib/systemd/system/nginx.service ##create profile [Unit] Description=nginx ##describe After=network.target ##Describe service type [Service] Type=forking ##Background operation form PIDFile=/usr/local/nginx/logs/nginx.pid ##PID file location ExecStart=/usr/local/nginx/sbin/nginx ##Startup service ExecReload=/usr/bin/kill -s HUP $MAINPID ##Overload configuration according to PID ExecStop=/usr/bin/kill -s QUIT $MAINPID ##Terminate process according to PID PrivateTmp=true [Install] WantedBy=multi-user.target [root@localhost ~]# chmod 754 /lib/systemd/system/nginx.service ##Set execution permission [root@localhost ~]# systemctl stop nginx.service ##Close nginx [root@localhost ~]# systemctl start nginx.service ##open
8. Write a log splitting script
[root@localhost nginx-1.12.0]# vim fenge.sh
#!/bin/bash
#Filename:fengge.sh ##Descriptive information
d=$(date -d "-1 day" "+%Y%m%d" ) ##Display the time one day before the system and generate a date string, such as "November 11, 2019"
logs_path="/var/log/nginx" ##Storage path after log segmentation
pid_path="/usr/local/nginx/logs/nginx.pid" ##The process number file of Nginx
[ -d $logs_path ] || mkdir -p $logs_path ##Determine whether there is a log split storage path. If not, create the path
mv /usr/local/nginx/logs/access.log ${logs_path}/test.com-access.log-$d
##Move out the access log under the original path and generate it to the created path, and generate a log file named after the date
kill -USR1 $(cat $pid_path) ##End the previous process number to generate a new process number
find $logs_path -mtime +30 | xargs rm -rf ##Find and delete the file 30 days ago under the path. xargs is used to take the processing result of the previous command as the parameter of the command after the pipe symbol
[root@localhost nginx-1.12.0]# chmod +x fenge.sh ##Grant script execution permission
[root@localhost nginx-1.12.0]# ./ fenge.sh
[root@localhost nginx-1.12.0]# ls /var/log/nginx
test.com-access.log-20191112 ##View the log partition file generated under the specified path
[root@localhost nginx-1.12.0]# ls /usr/local/nginx/logs
access.log error.log nginx.pid ##Check the Nginx log directory, and another access.log is generated automatically