system

tony2099 included in system

2021-05-01 336 words 2 minutes

Contents

system Understanding Systemd Units and Unit Files

what

is the init process(pid=1); a replacement of sysV and upstart;

初始化系统：start the rest of system
管理服务/进程

the unit(resouce) type

service: an applicaton;
target: g group of unit;
timer: like cron
socket: the servie active by a socket

the process of a application

start: ExecStart
stop:
1. use ExecStop=
2. send signTerm, after 90s, send SINGKILL

configure file

file path:

`/usr/lib/systemd/system/: install by package manager
/etc/systemd/system/ : user local create

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11


[Unit]
Description=<description about this service>

[Service]
User=<user e.g. root>
WorkingDirectory=<directory_of_script e.g. /root>
ExecStart=<script which needs to be executed>
Restart=always

[Install]
WantedBy=multi-user.target

1
2


systemctl list-unit-files # list install units
systemctl list-unit   # list unit in memory

the base resoruce systemed manage

unit file example

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13


[Unit]
Requires=network.target remote-fs.target
After=network.target remote-fs.target

[Service]
Type=simple
User=kafka
ExecStart=/home/kafka/kafka/bin/zookeeper-server-start.sh /home/kafka/kafka/config/zookeeper.properties
ExecStop=/home/kafka/kafka/bin/zookeeper-server-stop.sh
Restart=on-abnormal

[Install]
WantedBy=multi-user.target

syntax

unit

required: activate the target
after: make sure that current unit be started after target unit is active;

command

systemctl list-units: list all install files
1. –type: unit type
2. –state: load, acive, enabled
systemctl list-units --type=service --state=active

components

what: view log of systemd service

collect type:

auto collect: standout, standerr

upload to syslog server :

1
2
3
4
5



server {
   error_log syslog:server=unix:/dev/log;
   access_log syslog:server=unix:/dev/log;
}

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30




command： 
journalctl -f -u nginx 
journalctl -p err -u nginx



## the best practice to start a service

1. create user
2. assign the user privilege of some files
3. create unit files

## run level

 define a mode  a os run in  

1. shut down
2. single user
3. multi-user
5. multi-user with gui
6. reboot the system

systemd use target to descirpt it ;

target have two meaning: in a state and  a set of service;

1. network.target：
2. network-online.targe