Advantages of systemd over sysVinit



I think the reason Systemd strikes such a nerve is the way it was introduced in the first place. I have no real problem with it. It works fine on the Ditros I've used that use it.
I'll just say this; trying to explain the pros and cons of the various init systems to a fairly recent Linux convert would be like trying to explain the color blue to someone born blind. They've only ever known systemd.....so there's NO WAY they could possibly understand (since systemd will be all they've used).

Poettering's just a prat.....and a very self-centered one at that.


Mike. o_O
That I agree with, I'm so much used to systemd and like it very much, but otherwise don't know much about init.
 
^^^.....that from @CaffeineAddict

I know close to zero re init
probably not much more re systemd

What I DO know is......IT WORKS.

I have a PC which brings a smile to my face, every day. That is no mean feat.

Onward & Upward, LINUX.
 
I get many of the irks of Systemd, but I will say this. To move forward, sometimes you first must take a step back. Systemd set the ground work for many things to move forward. Sure, I've long used SysV and it worked really well, but it had real limitations to it also. Systemd does a lot to improve things, but obviously. You will always have SysV purists who will complain. They are more than welcome to go back to Pay phones. hah I will keep my smartphone. ;)

Systemd has a lot to offer. Though I wish someone could have slapped some sense into Lennart Poettering. Making completely unnecessary long ass commands like "systemctl list-unit-files" deserves good solid punch to the throat.
 
So does systemvinit.
systemD brought more standardization to the Linux distributions that use it. I remember the times with systemV you would have one init script to startup a service and then on another distribution another init script to start the same service. Before if I wanted to automatically startup a service a boot I would have to write an init script, now if I want to do the same I just write a unit file of several lines and most of the time easier than having to write an init script. And you can still call on init script in those unit files if you need them.

Everyone that hasn't read @osprey reply should do so, here's part of it.
Also I'm tired having to hear the comparison to Windows when it comes to Linux components, ie: systemD en KDE, etc. They are different than not the same, just like a fish and a birds aren't the same or a human and martian, and if does bother you so much just switch to BSD or DOS.
 
Last edited:
The guy who wrote "Biggest Myths" is Mr. Pottering. Of course he's going to say, "hey, systemd is great." Do you really think he's going to say it sucks? :rolleyes:

Although not specifically about systemd itself, I think this guy is right on the money:
https://dev1galaxy.org/viewtopic.php?pid=49278#p49278

Beware of bloatware, it's dangerous (seriously).
 
Last edited:
Beware of bloatware, it's dangerous (seriously).
Then the both the Linux kernel and systemD are bloat, because the Linux kernel has more lines of code than systemD.
Code:
systemD has lines of code: 1107244
Linux has lines of code: 27318023
Only Xserver and Wayland don't seem to be bloat because they have less than 1 million lines of code.
Code:
XServer has lines of code: 414086
Wayland has lines of code:  30903
I guess we should all switch to the Hurd kernel since that seems the also not be bloat because it has less than 1 million lines of code. ;)
Code:
Hurd has lines of code: 227986
 
Just joking around! ;)

From CoPilot.

Let’s delve into some of the risks and downsides associated with using systemd:
  1. Complexity and Bloat:
  2. Monoculture Concerns:
  3. Security Implications:
  4. Dependency on Privileges:
    • Ironically, security often requires privileges. For example, services may need root privileges to set up custom mount namespaces or open low-numbered ports.
    • While the service manager (like systemd) runs with high privileges, services themselves shouldn’t. The hardening setup often necessitates elevated privileges 3.
  5. Resistance from Traditionalists:
In summary, while systemd offers powerful features and benefits, it’s essential to weigh these against the potential risks and consider alternatives based on specific use cases and preferences.
Thanks for actually explaining, most anti-systemd people don't.
 
That's not true.
You're not getting the point. The word bloat seems to be overused now days and everyone seems to have a different definition of bloat when talking about bloat. The word bloat is used so much on Forums and Reddit that the word bloat has become more of joke and can't be taking seriously anymore most of the time. Sure systemD has downsides, but so does every other init system nothing is perfect.
 
Last edited:
I guess we should all switch to the Hurd kernel since that seems the also not be bloat because it has less than 1 million lines of code.
Sorry but your logic does not hold.

I used to criticize code which has lots of LOC and was happy to see I can do "the same" with twice as less code.
Up until one day a coder told me I'm wrong because more code means more tests done which in turn means handling more cases like security patches etc which in turn costs to write more code.

I only then realized my programs still have a lot of work more to do to achieve adequate stability comparable to finished software with more LOC.
 
Sorry but your logic does not hold.
If you read my last reply, I was trying to get to the point across showing that everyone calls everything bloat now days and I was using lines of code just to have some sort of comparison which yes isn't realistic. It's hard to exaggerate something just in text and making it sound like a joke.
 
Last edited:
@ron.alan wrote:
The guy who wrote "Biggest Myths" is Mr. Pottering. Of course he's going to say, "hey, systemd is great." Do you really think he's going to say it sucks? :rolleyes:

Although not specifically about systemd itself, I think this guy is right on the money:

Beware of bloatware, it's dangerous (seriously).

Ad hominem arguments were always illogical, fallacious and of no merit.

Bloat is not a quantitative measure, but rather a measure of the wasteful use of resources. The accusation of bloat against systemd needs to be substantiated by reference to where code has been unnecessarily wasteful and not achieving some design purpose it set out to solve.

It's unclear why the link in post #20 was provided since the author on that site states:
This is not about diversity, or even systemd specifically. It's about dependency bloat.
and more generally about library dependencies which involve multiple parts of systems, including systemd, so it's a more general argument being proposed.

In relation to the link in post #33, there are too many false assertions in that link for the reader to have any confidence in the validity of the text. In relation to some of the major assertions: it's not the case that systemd "prevents the skipping of fsck while booting", nor is it the case that systemd "Disables Linux Magic SysRq Key", nor that systemd "doing DNS resolves wrong", nor that starting a filename with a hyphen "causes all sorts of problems", nor that "systemd takes over logging." There are also a number of opinions expressed, which may or may not have merit, but there is no corroborating explanation for the reader to make any assessment of the matters from that text.
 
Well, I'll leave it at this. If the xz backdoor had made it into the wild, most of you with systemd would have been exposed. I would not have been. For me, that's good enough to stay away from the-d.
 
Last edited:
Well, I'll leave it at this. If the xz backdoor had made it into the wild, most of you with systemd would have been exposed. I would not have been. For me, that's good enough to stay away from the-d.
lol. Do you have a compressed kernel installed on your Linux system? Of course you do. There is a good damn chance that xz compressed it and xz was on your system when this occurred.

Besides, you do realize xz is a secondary utility used by many tools that you likely have installed on your system. Most Linux systems have xz installed, because it's a "helper" tool.

This is like arguing that Mac / Linux is safer than Windows. To hell it is. Anyone who says that either doesn't understand the problem or is straight talking out of their ass.

I will let you figure out which it is as I'm not looking to argue. I'm just pointing out facts. The fact is xz isn't a systemd issue. It's a code management issue. SysV (or whatever init system you use) can just as easily be compromised by a "helper" tool because all Linux systems rely on these helper tools because remember. Linux is only the kernel. The kernel itself. Is NOT an operating system.
 
Last edited:

Members online


Top