1
00:00:00,000 --> 00:00:11,360
The basics of systemd consists of managing the state of a systemd unit.

2
00:00:11,360 --> 00:00:12,600
So how does that work?

3
00:00:12,600 --> 00:00:18,120
Well, you would use the systemctl command to manage everything that is going on.

4
00:00:18,120 --> 00:00:20,760
Systemctl start will start the unit.

5
00:00:20,760 --> 00:00:26,660
Systemctl status will show unit status, and systemctl stop will stop it.

6
00:00:26,660 --> 00:00:31,719
To find out about different unit properties, use systemctl show.

7
00:00:31,719 --> 00:00:35,720
Also important is to make sure that units are started automatically when your system

8
00:00:35,720 --> 00:00:36,720
boots.

9
00:00:36,720 --> 00:00:40,040
We call that enabling and disabling units.

10
00:00:40,040 --> 00:00:46,020
Use systemctl enable with the optional argument minus minus now to start it now also if it

11
00:00:46,020 --> 00:00:51,340
hasn't been started yet, and that will make sure that the unit is automatically started.

12
00:00:51,340 --> 00:00:57,759
And systemctl disable minus minus now takes away automatic starting of the unit.

13
00:00:57,759 --> 00:01:03,840
Systemctl list units will show all units currently loaded, and systemctl list dependencies shows

14
00:01:03,840 --> 00:01:07,879
all that was loaded to get to the current system state.

15
00:01:07,879 --> 00:01:13,400
And finally, the command systemctl list unit files shows all unit files on this system,

16
00:01:13,400 --> 00:01:15,360
and that can be convenient as well.

17
00:01:15,360 --> 00:01:17,040
Let's check this out.

18
00:01:17,040 --> 00:01:21,040
So I'm going to demonstrate the HHPD unit.

19
00:01:21,739 --> 00:01:26,379
To start it, I need to install it using dnf install minus y hhtpd.

20
00:01:26,379 --> 00:01:32,620
I like to demonstrate HHPD because it's not an essential service that will break functionality

21
00:01:32,620 --> 00:01:35,059
if something is going wrong.

22
00:01:35,059 --> 00:01:42,260
Now let's start by using systemctl status on HHPD where we can see that systemd is aware

23
00:01:42,260 --> 00:01:49,220
of the HHPD.service, but the white dot here is indicating that currently it's inactive.

24
00:01:49,220 --> 00:01:55,540
We can also see that systemd has found the unit file, and use ellipse systemd system,

25
00:01:55,540 --> 00:02:01,160
and the name is hhtpd.service, but currently it's disabled and the preset is disabled.

26
00:02:01,160 --> 00:02:08,899
Let me change that using systemctl enable minus minus now hhtpd, which is not only going

27
00:02:08,899 --> 00:02:14,080
to flag it for automatic start, but which will start it right now as well.

28
00:02:14,080 --> 00:02:19,679
So when you enable a unit, it flags for automatic start, and behind the scenes, a symbolic link

29
00:02:19,679 --> 00:02:25,699
is created in the etc systemd system multi-user target dot once directory.

30
00:02:25,699 --> 00:02:30,320
And this symbolic link means that if you enter the multi-user target, it will follow the

31
00:02:30,320 --> 00:02:34,960
symbolic link, and as a result, it will automatically start the unit.

32
00:02:34,960 --> 00:02:37,240
Now let's investigate the status again.

33
00:02:37,240 --> 00:02:40,240
As you can see, there's a lot more information now.

34
00:02:40,240 --> 00:02:45,460
So we can see that it is enabled, preset disabled, that means that the default state

35
00:02:45,460 --> 00:02:49,800
of the unit as defined in the package is disabled.

36
00:02:49,800 --> 00:02:55,139
The unit is active and running, and we can even see the main process ID of the unit.

37
00:02:55,139 --> 00:02:57,779
That's convenient for process management.

38
00:02:57,779 --> 00:03:02,619
There is the number of tasks, there is the amount of memory it is using, and more relevant

39
00:03:02,619 --> 00:03:07,179
information, including the logs for this unit.

40
00:03:07,179 --> 00:03:11,259
These logs are convenient because if anything doesn't work out well, well, you can easily

41
00:03:11,259 --> 00:03:14,179
troubleshoot right here.

42
00:03:14,179 --> 00:03:17,520
Now the unit works with different parameters.

43
00:03:17,520 --> 00:03:22,539
We call them directives, and if you want to know more about these directives, use systemctl

44
00:03:22,539 --> 00:03:26,460
show on httpd.service.

45
00:03:26,460 --> 00:03:31,220
So systemctl show followed by the name of the unit shows all of the directives that

46
00:03:31,220 --> 00:03:35,300
are relevant for this specific unit.

47
00:03:35,300 --> 00:03:38,940
And there you can see many directives, and sometimes you will wonder, what is this directive

48
00:03:38,940 --> 00:03:39,940
doing?

49
00:03:39,940 --> 00:03:45,100
Well, if ever you are wondering what is this directive doing, you can look that up.

50
00:03:45,100 --> 00:03:48,580
Let's look up IP accounting, for instance.

51
00:03:48,580 --> 00:03:53,860
To do so, I'm using man systemd.directives.

52
00:03:53,860 --> 00:03:57,179
And there you can find the list of all directives that exist.

53
00:03:57,179 --> 00:04:04,619
So when I search for IP accounting using forward slash IP account, there we can see

54
00:04:04,619 --> 00:04:09,740
that IP accounting is further defined in systemd resource control 5.

55
00:04:09,740 --> 00:04:10,979
That's a man page.

56
00:04:10,979 --> 00:04:22,339
So if I want to know more about IP accounting, I am going to use man systemd.resourcecontrol.

57
00:04:22,339 --> 00:04:27,100
And there I can look it up, IP accounting, and I get a description of this directive.

58
00:04:27,100 --> 00:04:29,220
Now this was just one example.

59
00:04:29,220 --> 00:04:32,299
It will work for all directives.

60
00:04:32,299 --> 00:04:36,019
Now we have a couple of units that are listed right now.

61
00:04:36,019 --> 00:04:41,779
I like to figure out which these are, and to do so, there is systemctl list units.

62
00:04:41,779 --> 00:04:46,739
That will show you a list of all the units and their current state.

63
00:04:46,739 --> 00:04:51,459
That can be convenient, but this is focused on units that are currently active.

64
00:04:51,459 --> 00:04:57,700
If you want to see units that are inactive as well, I would suggest you use systemctl

65
00:04:57,700 --> 00:04:59,339
list unit files.

66
00:04:59,339 --> 00:05:04,980
There we can see the name of the units, nicely grouped by the unit type and the current state

67
00:05:04,980 --> 00:05:05,980
and the preset.

68
00:05:05,980 --> 00:05:12,100
So you can see if it is disabled or enabled, and that might give you a hint on what you

69
00:05:12,100 --> 00:05:16,980
need to do to get your system in the desired state.

70
00:05:16,980 --> 00:05:21,660
Now for a nice overview of everything that is happening in the current situation, you

71
00:05:21,660 --> 00:05:28,480
can also use systemctl list dependencies.

72
00:05:28,480 --> 00:05:32,739
That is showing the current state where you are, so the default target.

73
00:05:32,739 --> 00:05:36,420
Default target on a graphical system is graphical.target.

74
00:05:36,420 --> 00:05:39,700
And in the graphical.target, we have a couple of units.

75
00:05:39,700 --> 00:05:44,700
Then graphical.target includes the multi-user target, which has units as well.

76
00:05:44,700 --> 00:05:48,579
And you can see units that are currently running, but also units that are currently

77
00:05:48,579 --> 00:05:50,179
not running.

78
00:05:50,179 --> 00:05:55,100
And even in some cases with the Red Cross units that have had an issue.

79
00:05:55,100 --> 00:06:00,339
I like systemctl list dependencies a lot, because you get an overview of everything

80
00:06:00,339 --> 00:06:04,100
and you see the relation between the units in an easy way.