Linux systemd 基础及命令

From the project web page:

systemd is a suite of basic building blocks for a Linux system. It provides a system and service manager that runs as PID 1 and starts the rest of the system. systemd provides aggressive parallelization capabilities, uses socket and D-Bus activation for starting services, offers on-demand starting of daemons, keeps track of processes using Linux control groups, maintains mount and automount points, and implements an elaborate transactional dependency-based service control logic. systemd supports SysV and LSB init scripts and works as a replacement for sysvinit. Other parts include a logging daemon, utilities to control basic system configuration like the hostname, date, locale, maintain a list of logged-in users and running containers and virtual machines, system accounts, runtime directories and settings, and daemons to manage simple network configuration, network time synchronization, log forwarding, and name resolution.

支持的主要单元：服务单元(.service)，挂载单元(.mount)，目标单元(.target)等

默认情况下，systemd会在启动单元的同时启动其所有的Requires和Wants依赖组件。
Requires和Wants定义依赖关系并没有设定顺序，使用顺序关键字(Before和After)来设定顺序
        Before：当前单元会在Before中列出的单元之前启动
        After： 当前单元在After中列出的单元之后启动

记住这个原则即可：
        不要更改系统单元目录，因为它由系统来维护。可以在系统配置目录中保存你的自定义设置。在选择更改/usr还是更改/etc时，永远选择/etc

# There are two safe ways to modify a unit without touching the original file: 
# create a new unit file which overrides the original unit or create drop-in snippets which are applied on top of the original unit.
# Replace or Drop-in unit files, with parameter --full is Replace mode, otherwise Drop-in mode
$ sudo systemctl edit [--full] ssh.service
	
# To revert any changes to a unit made using systemctl edit
$ systemctl revert ssh.service

默认情况下，systemctl edit使用nano作为编辑器，使用如下方法改变使用vi作为编辑器

$ vi ~/.bashrc
    export SYSTEMD_EDITOR=vim
$ sudo visudo
	Defaults  env_keep += "SYSTEMD_EDITOR"

Install节

sshd.service
[Unit]
Description=OpenSSH server daemon
After=syslog.target network.target auditd.service
	
[Service]
EnvironmentFile=/etc/sysconfig/sshd
ExecStartPre=/usr/sbin/sshd-keygen
ExecStart=/usr/sbin/sshd -D $OPTIONS
ExecReload=/bin/kill -HUP $MAINPID
	
[Install]
WantedBy=multi-user.target

单元文件中的[Install]区块很重要，因为它告诉我们怎样使用systemd的WantedBy和RequiredBy依赖关系
    这部分配置的目标模块通常是特定运行目标的.target文件，用来使得服务在系统启动时自动运行。这个区段可以包含三种启动约束：
        WantedBy：和 Unit 段的 Wants 作用相似，只有后面列出的不是服务所依赖的模块，而是依赖当前服务的模块。
            它的值是一个或多个 Target，当前 Unit 激活时（enable）符号链接会放入 /etc/systemd/system目录下面以+.wants后缀构成的子目录中，如 "/etc/systemd/system/multi-user.target.wants/"
        RequiredBy：和 Unit 段的 Requires作用相似，只有后面列出的不是服务所依赖的模块，而是依赖当前服务的模块。
            它的值是一个或多个 Target，当前 Unit 激活时，符号链接会放入 /etc/systemd/system目录下面以+.required后缀构成的子目录中
        Also：当前 Unit enable/disable 时，同时 enable/disable 的其他 Unit
        Alias：当前 Unit 可用于启动的别名

Unit 模板文件的写法与普通的服务 Unit 文件基本相同，不过 Unit 模板的文件名是以 @ 符号结尾的。通过模板启动服务实例时，需要在其文件名的 @ 字符后面附加一个参数字符串。
    Systemd 在运行服务时，总是会先尝试找到一个完整匹配的 Unit 文件，如果没有找到，才会尝试选择匹配模板。

Unit状态：
    enabled：已建立启动链接
    disabled：没建立启动链接
    static：该配置文件没有 [Install] 部分（无法执行），只能作为其他配置文件的依赖
    masked：该配置文件被禁止建立启动链接

systemd启动流程

 

target配置流程(实际启动时从最底层的local-fs-pre.target开始执行)
      

 -> default.target (=graphical.target by default) 
        Requires=multi-user.target
        Wants=display-manager.service
        Conflicts=rescue.service rescue.target
        After=multi-user.target rescue.service rescue.target display-manager.service

            This tells systemd to start everything in the multi-user.target before starting the graphical target. 
            Once that's done, the "Wants" entry tells systemd to start the display-manager.service service (/etc/systemd/system/display-manager.service), which runs the GNOME display manager (/usr/sbin/gdm).
      

 -> multi-user.target
        Requires=basic.target
        Conflicts=rescue.service rescue.target
        After=basic.target rescue.service rescue.target

            This tells systemd to start everything in the /lib/systemd/system/basic.target target before starting the other multi-user services. 
            After that, for the multi-user.target, all units (services, targets, etc.) in the /etc/systemd/system/multi-user.target.wants and /lib/systemd/system/multi-user.target.wants directories are started. 
            When you enable a service, a symbolic link is placed in the /etc/systemd/system/multi-user.target.wants directory. 
            That directory is where you will find links to most of the services you think of as starting in multi-user mode (printing, cron, auditing, SSH, and so on).
      

 -> basic.target 
        Requires=sysinit.target
        Wants=sockets.target timers.target paths.target slices.target
        After=sysinit.target sockets.target paths.target slices.target tmp.mount

            This points systemd to the /lib/systemd/system/sysinit.target, which must start before the basic.target can continue. 
            The basic.target target file starts the firewalld and microcode services from the /etc/systemd/system/basic.target.wants directory and services for SELinux, kernel messages, and loading modules from the /lib/systemd/system/basic.target.wants directory.

-> sysinit.target 
        Conflicts=emergency.service emergency.target
        Wants=local-fs.target swap.target
        After=local-fs.target swap.target emergency.service emergency.target

            Besides mounting file systems and enabling swap devices, the sysinit.target starts targets, services, and mounts based on units contained in the /lib/systemd/system/sysinit.target.wants directory. 
            These units enable logging, set kernel options, start the udevd daemon to detect hardware, and allow file system decryption, among other things. 
The /etc/systemd/system/sysinit.target.wants directory contains services that start iSCSI, multipath, LVM monitoring and RAID services.

-> local-fs.target
        Conflicts=shutdown.target
        After=local-fs-pre.target

            There are no services associated with the local-fs-pre.target target (you could add some to a "wants" directory if you like). 
            However, units in the /lib/systemd/system/local-fs.target.wants directory import the network configuration from the initramfs, run a file system check (fsck) on the root file system when necessary, and remounting the root file system (and special kernel file systems) based on the contents of the /etc/fstab file.

启动时通过/etc/systemd/system/unit.target.wants目录来决定哪些服务会被当前target启动

systemctl service和target相关命令

# 获取当前正在使用的运行目标
# systemctl list-units --type=target
	
# 只重新加载unit单元的配置
$ systemctl reload unit：
# 重新加载所有的单元配置
$ systemctl daemon-reload：
	
# 显示单元的配置文件或配置参数
$ systemctl [cat|show] docker.service

# 显示单元的依赖关系
$ systemctl list-dependencies docker.service
	
# 查看当前系统的所有 Target
$ systemctl list-unit-files --type=target

# 查看一个 Target 包含的所有 Unit
$ systemctl list-dependencies multi-user.target

# 查看启动时的默认 Target
$ systemctl get-default

# 设置启动时的默认 Target
$ sudo systemctl set-default multi-user.target

# 切换 Target 时，默认不关闭前一个 Target 启动的进程，systemctl isolate 命令改变这种行为，关闭前一个 Target 里面所有不属于后一个 Target 的进程
# This will only change the current target, and has no effect on the next boot
$ sudo systemctl isolate multi-user.target

 journalctl命令

# 查看所有日志（默认情况下 ，只保存本次启动的日志）
$ sudo journalctl

# 查看内核日志（不显示应用日志）：--dmesg 或 -k
$ sudo journalctl -k

# 查看系统本次启动的日志（其中包括了内核日志和各类系统服务的控制台输出）：--system 或 -b
$ sudo journalctl -b
$ sudo journalctl -b -0

# 查看上一次启动的日志（需更改设置）
$ sudo journalctl -b -1

# 查看指定服务的日志：--unit 或 -u
$ sudo journalctl -u docker.servcie

# 查看指定服务的日志
$ sudo journalctl /usr/lib/systemd/systemd

# 实时滚动显示最新日志
$ sudo journalctl -f

# 查看指定时间的日志
$ sudo journalctl --since="2012-10-30 18:17:16"
$ sudo journalctl --since "20 min ago"
$ sudo journalctl --since yesterday
$ sudo journalctl --since "2015-01-10" --until "2015-01-11 03:00"
$ sudo journalctl --since 09:00 --until "1 hour ago"

# 显示尾部的最新 10 行日志：--lines 或 -n
$ sudo journalctl -n

# 显示尾部指定行数的日志
$ sudo journalctl -n 20

# 将最新的日志显示在前面
$ sudo journalctl -r -u docker.service

# 改变输出的格式：--output 或 -o
$ sudo journalctl -r -u docker.service -o json-pretty

# 查看指定进程的日志
$ sudo journalctl _PID=1

# 查看某个路径的脚本的日志
$ sudo journalctl /usr/bin/bash

# 查看指定用户的日志
$ sudo journalctl _UID=33 --since today

# 查看某个 Unit 的日志
$ sudo journalctl -u nginx.service
$ sudo journalctl -u nginx.service --since today

# 实时滚动显示某个 Unit 的最新日志
$ sudo journalctl -u nginx.service -f

# 合并显示多个 Unit 的日志
$ journalctl -u nginx.service -u php-fpm.service --since today

# 查看指定优先级（及其以上级别）的日志，共有 8 级
# 0: emerg
# 1: alert
# 2: crit
# 3: err
# 4: warning
# 5: notice
# 6: info
# 7: debug
$ sudo journalctl -p err -b

# 日志默认分页输出，--no-pager 改为正常的标准输出
$ sudo journalctl --no-pager

# 以 JSON 格式（单行）输出
$ sudo journalctl -b -u nginx.service -o json

# 以 JSON 格式（多行）输出，可读性更好
$ sudo journalctl -b -u nginx.serviceqq
 -o json-pretty

# 显示日志占据的硬盘空间
$ sudo journalctl --disk-usage

# 指定日志文件占据的最大空间
$ sudo journalctl --vacuum-size=1G

# 指定日志文件保存多久
$ sudo journalctl --vacuum-time=1years

参考：

https://cloud.tencent.com/developer/article/1516125

https://wiki.archlinux.org/title/Systemd#Running_services_after_the_network_is_up

https://linux.cn/article-5457-1.html

https://access.redhat.com/articles/754933