linux system services

1. System startup process

2. Start operation level

2.1 what is the operation level

Run level: refers to the functional level at which the operating system is running

system V init run level	systemd target name	effect
0	runlevel0.target,poweroff.target	Shut down
1	runlevel1.target,rescue.target	Single user mode (change password)
2	runlevel2.target,multi-user.target
3	runlevel3.target,multi-user.target	Character interface (maximum)
4	runlevel4.target,multi-user.target
5	runlevel5.target,graphical.target	Graphical interface
6	runlevel6.target,reboot.target	restart

2.2 adjust operation level

systemd uses targets instead of runlevels. There are two
Main objectives:
. multi-user.target: similar to runlevel 3
. graph ical.target: similar to runlevel 5
1. View the default operation level of the system

[root@localhost ~]# systemctl get-default 
multi-user.target

2. To set the default target, run

[root@localhost ~]# systemctl set-default TARGET.target
# TARGET.target run target

3.systemd management

3.1 origin of SYSTEMd

Linux has always adopted init process as its ancestor process, but
Yes, init has two disadvantages:
1. The system takes a long time to start. The init process is started serially and only
After the previous process is started, the next process will be started;
2. The startup script is complex. After initializing the system, many scripts will be loaded. The scripts will deal with their own situations, and there are many and miscellaneous scripts;
3. Write script;
systemd is the system daemon. systemd was born to solve the above problems
The goal of systemd is to provide a complete solution for system startup and management;
Centos5 is slow to start, and the process of serial start, regardless of whether the processes depend on each other or not.
The startup speed of Centos6 has been improved. Processes with dependencies are started in turn, while others without dependencies are started in parallel and synchronously.
All processes in Centos7 are started in parallel whether they have dependencies or not (of course, most of the time, the process is not really started, but only a signal or mark, which will be really started when it is really used.)

3.2 advantages of SYSTEMd

1. The latest systems use systemd to manage RedHat7, CentOS7 and Ubuntu 15;
2. Centos7 supports parallel startup service, which significantly improves startup efficiency;
3. When Centos7 is shut down, only running services are shut down, while Centos6 is shut down once;
4. The start and stop of Centos7 services are not managed by scripts, that is, / etc / init There is no script under D;
5. Centos7 uses systemd to solve the defects of the original mode. For example, the original service will not close the subprocess generated by the program;

3.3 systemd related commands

/usr/lib/systemd/system /: service startup and shutdown files. Start, stop and reload the files through systemctl command

command	meaning
systemctl start crond	Start service
systemctl stop crond	Out of Service
systemctl restart crond	Restart service
systemctl reload crond	Overload service
systemctl enable crond	The service is set to start and run
systemctl disabled crond	The service is set to start and not run
systemctl daemon-reload crond	Creating a service startup file requires overloading the configuration
systemctl list-unit-files	View the startup and disabling of services at various levels
systemctl is-enabled crond.service	Check whether a specific service is self started
systemctl is-active crond	Check whether the service is running

3.4 systemd management Nginx

1. Compiling nginx

[root@localhost nginx-1.21.1]# wget http://nginx.org/download/nginx- 1.21.1.tar.gz
[root@localhost nginx-1.21.1]# tar xf nginx-1.21.1.tar.gz 
[root@localhost nginx-1.21.1]# ./configure --prefix=/usr/local/nginx-1.21.1 --with-http_ssl_module --with-http_stub_status_module --without-http
[root@localhost nginx-1.21.1]# make && make install
[root@localhost nginx-1.21.1]# ln -s /usr/local/nginx-1.21.1/ /usr/local/nginx

2. Conventional startup nginx mode

4. Rescue mode

4.1 scenario 1 - forget the super administrator password

How to use single user mode to change system password? Take Centos7 system as an example:
. Step 1: when the Linux system host is restarted and the boot interface appears, press the e key on the keyboard to enter the kernel editing interface
. Step 2: add forcing = 0 init = / bin / bash after the line of Linux 16, and then press Ctrl + x to enter
. Step 3: enter the single user mode of the system, enter the following commands in sequence, restart the operating system and log in with a new password

1. mount -o rw,remount / default / partition read only,
Remount to read-write mode
2. echo "123" | passwd --stdin root: not
Interactive password modification
3. exec /sbin/init: reboot the system

4.2 scenario 2 - data backup required due to system damage

When the system breaks down and can't log in to the system, but you need to copy the data inside, what should I do?
Step 1: Mount ISO image file, modify BIOS and adjust
The DVD disc is the first boot;
Step 2: select Troubleshooting and continue to select
Rescue a CentOS system rescue mode;
Step 3: enter 1, and then execute the command chroot /mnt/sysimage to mount the real system;
Step 4: back up the data files in the system to other disks;

4.3 scenario 3 - how to repair a grub file accidentally deleted by Centos

Step 1: simulate the false deletion fault RM -rf / boot / Grub2 & & reboot
Step 2: the system cannot be started normally (prompt grub cannot be found)
To)
Step 3: then enter the rescue mode according to the previous operation and execute
chroot /mnt/sysimage mounts the real operating system;
Step 4: use grub2 related commands to repair
Grub2 install / dev / SDA re add mbr boot
grub2-mkconfig -o /boot/grub2/grub.cfg
Rebuild configuration
Enter exit
reboot again
ls /boot/grub2/grub.cfg

After restart

5. System optimization

cpu:
. 1. Virtualization support
. 2. cpu hyper threading
disk
. 1,SAS
. 2,SSD

5.1 adjusting the source

The yum warehouse provided by the default installation system is a foreign site. You can modify the site to a domestic site to speed up the download of software packages

# base 
[root@oldxu ~]# wget -O /etc/yum.repos.d/CentOS-Base.repo https://mirrors.aliyun.com/repo/Centos- 7.repo 
# epel 
[root@oldxu ~]# wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-7.repo 
# mysql, zabbix, elk packages

5.2 turn off the firewall

By default, the firewall will be closed to avoid affecting the use of later services;
. Virtual machine: there are security group products provided by hardware to provide protection;
. Physical machine: generally, companies have hardware firewalls at the entrance level;

# firewalld 
[root@oldxu ~]# systemctl stop firewalld
[root@oldxu ~]# systemctl disabled firewalld 
# selinux 
[root@oldxu ~]# setenforce 0 [root@oldxu ~]# sed -i '/^SELINUX=/c SELINUX=disabled' /etc/selinux/config

5.3 ulimit resource constraints

The ulimit command can control system resources
. - u: Limit the maximum number of processes that ordinary users can open; (per user)
. - n: Limit the number of file descriptors that users can allocate;

5.3.1 limit the maximum number of processes

1. Limit the maximum number of open processes per user to 3

[root@localhost ~]# ulimit -u 500
#Temporarily effective, the current window switches users

5.3.2 limit the number of open files

1. Limit the maximum open file descriptor of the process to 10

[root@localhost ~]# ulimit -n 10
#Temporarily effective, the current window switches users

2. Write a python program to simulate opening multiple files

#!/usr/bin/env python 
import time,os
from threading import Thread

print(os.getpid())

def new_file(n):
    with open('%s.file' %n,mode='wt') as f1:
                time.sleep(2000)

if __name__ == "__main__":
        count=1
        while True:
                Thread(target=new_file,args= (count,)).start()
        count+=1

3. After waiting for a period of time, the program will prompt when the open file exceeds the limit
Abnormal;

Exception in thread Thread-16128:
Traceback (most recent call last):
  File "/usr/lib64/python2.7/threading.py", line 812, in __bootstrap_inner
  File "/usr/lib64/python2.7/threading.py", line 765, in run
  File "o.py", line 8, in new_file
IOError: [Errno 24] Too many open files: '1.file'

5.3.3 adjusting ulimit limit parameters

Adjusting the number of open files and processes through ulimit is a temporary operation, so it needs to be permanently configured
. Configuration file: / etc / security / limits conf
. Adjustment mode:
Soft: soft limit. If it exceeds, you will be prompted;
Hard: hard limit, if exceeded, stop;

[root@localhost ~]# tail /etc/security/limits.conf 
#ftp             hard    nproc           0
#@student        -       maxlogins       4

# End of file
# max user processes 
* soft nproc 60000 
* hard nproc 60000 
# open files # 
* soft nofile 100000 
#2 ^ 16-1 (the 16th power of 2 cuts 1)
* hard nofile 100000
#2 ^ 16-1 (the 16th power of 2 cuts 1)
# At the system level, add the following fields (only after adjusting the kernel can it take effect)
[root@localhost ~]# tail -10 /etc/sysctl.conf 
# /usr/lib/sysctl.d/, /run/sysctl.d/, and /etc/sysctl.d/.
#
# Vendors settings live in /usr/lib/sysctl.d/.
# To override a whole file, create a new file with the same in
# /etc/sysctl.d/ and put new settings there. To override
# only specific settings, add a file with a lexically later
# name in /etc/sysctl.d/ and put new settings there.
#
# For more information, see sysctl.conf(5) and sysctl.d(5).
fs.file-max = 100000

Keywords: Linux

Added by howler on Mon, 10 Jan 2022 19:08:47 +0200

Programming VIP