As long time SoylentNews community member Marand observed during some recent discussion of severe systemd boot problems, it turns out that systemd disables the magic SysRq key.
The magic SysReq key is described at Wikipedia as:
[...] a key combination understood by the Linux kernel, which allows the user to perform various low-level commands regardless of the system's state. It is often used to recover from freezes, or to reboot a computer without corrupting the filesystem.
A Fedora user who logged a bug report for this issue back in 2013 described the problem with systemd's unexpected and harmful default setting:
As systemd depends on many files on a rootfs, in case of any problems with rootfs, it is not able to do its basic function - control processes and (cleanly) shutdown/reboot when crtl-alt-del is pressed on local keyboard. As this is a feature, I'd like to ask to enable the sysrq by default on Fedora, otherwise it is not possible to reboot system even locally in case of emergency situation.
While that Fedora bug report is set to CLOSED NOTABUG, other Linux distros, like Mageia and Debian GNU/Linux, have restored the proper behavior.
Now that this problem has come to light, all Fedora users should evaluate whether or not they need to fix their systems to work around systemd's incorrect default setting. Users of other Linux distributions using systemd should also evaluate their systems, too, in case their distro has not yet fixed this unexpected bug.
Related Stories
A Debian user has recently discovered that systemd prevents the skipping of fsck while booting:
With init, skipping a scheduled fsck during boot was easy, you just pressed Ctrl+c, it was obvious! Today I was late for an online conference. I got home, turned on my computer, and systemd decided it was time to run fsck on my 1TB hard drive. Ok, I just skip it, right? Well, Ctrl+c does not work, ESC does not work, nothing seems to work. I Googled for an answer on my phone but nothing. So, is there a mysterious set of commands they came up with to skip an fsck or is it yet another flaw?
One user chimed in with a hack to work around the flaw, but it involved specifying an argument on the kernel command line. Another user described this so-called "fix" as being "Pretty damn inconvenient and un-discoverable", while yet another pointed out that the "fix" merely prevents "systemd from running fsck in the first place", and it "does not let you cancel a systemd-initiated boot-time fsck which is already in progress."
Further investigation showed that this is a known bug with systemd that was first reported in mid-2011, and remains unfixed as of late December 2014. At least one other user has also fallen victim to this bug.
How could a severe bug of this nature even happen in the first place? How can it remain unfixed over three years after it was first reported?
systemd ... (Score:0)
systemd, the prize in your cereal box.
viva relativity!!! (Score:1, Funny)
Soylentnews.org and long time user, viva relativity!!!
So to recap this weekend's systemd discoveries. (Score:4, Informative)
So this weekend we learned that:
1) Systemd prevents boot-time fsck runs from being canceled while in progress. [soylentnews.org]
2) Systemd (or the pid 1 part of it) can segfault, which is something that should never happen to pid 1. [soylentnews.org]
3) Systemd forces binary logs, but the log viewing utility doesn't display long lines properly. [soylentnews.org]
4) Systemd disables the Linux sysrq magic key. [soylentnews.org]
I couldn't believe 1). I wasn't surprised by 2) and 3). But 4) leaves me utterly shocked.
How can Debian justify pushing such obviously broken software into Debian 8, and thus onto all of its users?
Why aren't all of these problems throwing up red flags left and right within the Debian leadership and among the Debian maintainers?
Why isn't Debian 8 being delayed until systemd can be completely removed?
Re:So to recap this weekend's systemd discoveries. (Score:4, Interesting)
1) Systemd prevents boot-time fsck runs from being canceled while in progress.
According to one of the messages [debian.org] in the mailing list thread about the fsck problem, systemd actually prevents using ^C to break out of any in-progress task, not just fsck. Definitely bigger than just "deal with it it's only fsck". It seemed like a big deal, so I mentioned it in a follow-up comment in the fsck discussion.
It might not be hopeless, though; another person pointed out that when Debian first introduced parallel booting with sysv-init, they had to temporarily disable ^C in the same way because you couldn't reasonably guarantee the process you wanted to close was the one ^C was going to close. It's since been fixed and re-enabled, so maybe the same thing will happen with systemd. Of course, based on previous behaviour of systemd's leading people, it's just as likely to get closed as NOTABUG,STFU
Parent
CLOSED NOTABUG (Score:4, Informative)
How often do these idiots refuse to fix bugs? Seems like they pretty much won't admit that any shortcoming in their software exists. Like that time when they broke kernel logging and tried to say it wasn't a bug, and Linus had to scream at them and refuse to accept code by Kay until he fixed his shit (http://lkml.iu.edu/hypermail/linux/kernel/1404.0/01331.html).
I mean, even if you hit your head and think something like systemd is a good idea, you probably don't want software made by these guys.
novak
Just finished testing FreeBSD... (Score:1, Redundant)
We just did our first install of FreeBSD on a new server that is replacing an aging system and it went much more smoothly than any of us expected. There are certainly some differences between Linux and BSD but faced with having to learn the BSD way versus a bunch of hacks, workarounds, and commands relating to systemd, the choice is clear; we're fully embracing BSD.
To the systemd apologists who like to claim that nobody is forcing systemd on the rest of us and that there is an abundance of distros without it, all I can say is you need to wake the hell up. I am extremely sad with the current state of the Linux landscape and didn't want to leave it behind. However, we are a small company and we don't have the resources to waste on babysitting systemd or rolling our own distro; we need something that's boring and works -- like Debian used to... This choice you speak of is nearly non-existent and growing smaller as systemd sucks more of the ecosystem down with it. If you don't see the negative effect it's having then you're part of the problem. Linux projects need to stop reinventing shit over and over and they need to really listen to the users. Until then, FreeBSD is my light at the end of a very dark tunnel...
YASDDFB thread (Score:3, Insightful)
Yet Another SystemD Doesn't Fix Bugs Thread. It's pretty clear at this point the systemd developers have a take it or leave it approach to their code, don't bother reporting bugs. If the bug doesn't affect the systemd developers then it's irrelevant.
I run Kali linux on one of my laptops, dunno if it runs systemd or not (how can I tell?). But from what I've heard since last April, given the choice I'd be like Monty Python and "run away, run away!"
Systemd Free Linuxes (Score:3, Informative)
I often hear the complaint that users have no options besides systemd or BSD. Here is a quick list I got by browsing through distrowatch of linuxes which do not have systemd. I disallowed any distros based on Arch, Debian, and the like, as systemd is probably coming to them- by doing this I likely missed a few. Likewise, I used the page on distrowatch as my source of info on whether it runs systemd, so this list does not cover whether systemd is planned. Let me know if I made a mistake.
PCLinuxOS
Puppy Linux
Simplicity Linux
4MLinux
Slackware Linux
Gentoo Linux
Salix
Porteus
Alpine Linux
Tiny Core Linux
Absolute Linux
VectorLinux
LinuxConsole
SliTaz
Wifislax
Slackel
Zenwalk Linux
Pisi Linux
IPFire
CRUX
Legacy OS
AUSTRUMI
Quirky
GoboLinux
0linux
Kwort Linux
Devil-Linux
Slax
IPCop
Macpup
TouTou Linux
AgiliaLinux
Vine Linux
Thinstation
Turbolinux
Dragora GNU/Linux-Libre
Foresight Linux
openmamba GNU/Linux
Supurb Mini Server
Draco GNU/Linux
Ophcrack LiveCD
Plamo Linux
Burapha Linux Server
NuTyX
Zencafe GNU/Linux
Voidlinux
novak
Re:Systemd Free Linuxes (Score:4, Insightful)
Posting a list of alternate Linux distros that just happen to still be systemd-free right at this exact moment is completely missing the point. Besides the fact that a lot of these minor Linux distributions are made for narrow, specialized purposes, they are often based on a snapshot of a larger, more established distribution like Debian. They share "DNA" with one or more parent distributions, so to speak. As many have said, systemd is spreading out and rapidly infecting things exactly like a cancer. The next time any of those minor distros feel like taking a new snapshot of their base distro, they will become infected with systemd as well. And this is even before taking into account the inevitable encroachment of developers who have bought into systemd as being a good idea, who will soon start to only maintain the systemd-requiring versions of their software, thus forcing many minor "from-scratch" Linux distros to eventually contemplate adding systemd just to maintain some relevance in the overall Linux scene.
No, there is good reason why so many people are talking about completely walking away from Linux as a platform, because unlike other platforms like BSD, the Linux platform has failed to reject the very idea of what systemd is doing to the platform. Eventually I believe it will take monumental effort to avoid systemd anywhere in the Linux world, and people just don't feel like wasting that kind of time when they can just jump over to a BSD or even Windows and end up getting more productive work done. It is my considered opinion at this point that systemd is eventually going to end up basically being the death knell of Linux as a popular desktop/server platform. It may not become blindingly obvious until several years from now, but I believe that we will all look back and collectively agree that the advent of systemd was the inflection point in Linux's popularity. Unless of course Linux can somehow reverse course and divest itself of systemd completely and be reborn, but I don't foresee that happening. That's the only possible thing that will bring back any of the folks who are "jumping ship" right now to other platforms. But it would require a very deep and fundamental reevaluation of the core principles of the entire Linux platform.
¯\_ʕ◔.◔ʔ_/¯ LOL. I dunno. I'm just a bear.
... Peace out. Got bear stuff to do. 彡ʕ⌐■.■ʔ
Parent
Magic SysRq has helped me a lot (Score:3, Insightful)
Back when my X kept freezing randomly for seemingly no reason (it was a bad power supply, but that one was tough to figure out), I was using the magic SysRq daily. That prevented a whole lot of filesystem corruptiion. It's a good thing systemd wasn't there to interfere. I wouldn't have needed for systemd to add another problem to solve on top of my existing problem.
Maybe the reason I enabled the magic SysRq when I compiled the kernel was because I wanted it to be there for a reason. I don't need my init system silently disabling useful features, even if Lennart Poettering in his great wisdom thinks that ignorant users mashing every keyboard button at once and potentially accidentally rebooting their system are a serious problem. That's the same sort of user-contemptuous mindset that makes Gnome decide that the middle click paste is deprecated now even though it's the more efficient way to copy-paste.
Too bad devuan is a fucking scam. (Score:0)
Too bad devuan is a fucking scam where they took money and have done NO work.
They just talk about food and ban mysoginists and brooklyn jews and pat themselves on the back.
Fucking pieces of shit.
temporary measure, block apt installing systemd (Score:2, Informative)
The magic incantation to block apt from installing systemd when doing updates can be found at this link.
http://without-systemd.org/wiki/index.php/How_to_remove_systemd_from_a_Debian_jessie/sid_installation [without-systemd.org]
It currently works for me, though there is a reasonable chance of some new dependency stopping updates
from working in the future.
You guys are just thirsty (Score:0)
W/e, sysvinit had the system key disabled by default because it was seen as a security issue by quite a few big distros. You guys are just thirsty.
This doesn't seem like much of a story (Score:1)
I think we can all agree that whether SystemD is a good thing or not, it is designed to do init-type tasks.
The SystemD guys think that SystemD should legitimately be setting up the magic SysReq key. They picked a default value that "should be safe"... I don't know what that means but I guess it disables the magic SysReq key. I'll say it again: the default value disables the magic SysReq key.
Another guy thinks that the magic SysReq key is special, and SystemD shouldn't set it. Some other config file should set it. Further discussion of the bug showed that SystemD guys don't agree.
Okay, I'm trying to figure out why this is a top-level news story. If you have SystemD, you enable the magic SysReq key by editing a SystemD setup file. This is controversial... why?
IMHO if you are letting SystemD manage your system, it should be no big deal to enable the magic SysReq key by editing a SystemD config file rather than some other config file.
So, what have I missed here? Why is this a top-level story?
Can someone please (Score:0)
Submit this and/or the other systemd articles to Slashdot? As an upset Debian user I find these systemd articles a lot more important than the current drivel articles there right now, this stuff needs to get known.
alt-sysrq-REISUB (Score:1)
Retards Embrace Init System Utterly Broken
Encapsulated, Extend, Absorb (Score:0)
Has no-one else noticed exactly what systemd really is?