Slashdot Mirror
Systemd Adding Its Own Console To Linux Systems
An anonymous reader writes: The next version of systemd is poised to introduce an experimental "systemd-consoled" that serves as a user-space console daemon. The consoled furthers the Linux developers' goal of eventually deprecating the VT subsystem found within the Linux kernel in favor of a user-space driven terminal that supports better localization, increased security, and greater robustness of the kernel's seldom touched and hairy CONFIG_VT'ed code.
srsly?
And this is why I'm switching to OpenBSD. Fuck. This. Shit.
And fuck you Poettering. DIAF already.
Few people want systemd at all. Why it is being forced on us?
Please stop this madness!
Why is everyone so mad about it?
Is it really just me that has a shitload of problems with the current VT?
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion. -- Spazmania (174582)
All systemd needs now is an integrated web browser and a registry!
As long as I can still run vi in it, I'm good.
I'm just here for the entertainment. Now where are those villagers with pitchforks?
"A person is smart. People are dumb, panicky dangerous animals and you know it." - K
Does anyone really want "better localization" in terminals. My experience as a bilingual user from windows is that the less things are localized the better they work.
Making commands localized breaks script compatibility. (And that includes any output if that is parsed too.)
It has gone to the point where I get the English version of Windows rather than one adapted to my native language. The localization of some of the folder names makes things break and the translation of GUI elements obfuscates the function and makes it so that one has to translate everything to English and back to realize what the function is, especially when the original translator used every synonym for "device" he could possibly find.
Unless they have found a new revolutionary way to localize stuff that haven't been done before. Then it might actually work.
This is what is wrong with SystemD.... Do ONE job, do it well. Not replace the entire ecosystem.....
Sendmail is like emacs: A nice operating system, but missing an editor and a MTA.
Few people want systemd at all. Why it is being forced on us?
If you don't want to use all of Lennart Poettering's code, you can choose to use less.
Will this console run any worthwhile games? (Other than NetHack, that is.)
X-mas 2016 edition of SystemD:
- now with built in kitchen sink, linux kernel not needed.
--- Reality doesn't care about your opinions, it happens anyway and if you are in the way you'll get squished.
Linking to slides of a talk without the actual content of what's being presented doesn't give me the keys to understand why that change/new approach is necessary. I'm not against systemd or bashing it, I'm part of the people that don't understand half of what it's replacing to begin with.
Thirty seconds every six months on a system where the motherboard BIOS POST, each NIC firmware, the SCSI card firmware, spinning up the drives, and the RAID bios take around two to three minutes to complete. So not really much at all. I propose new features to systemd to parallelize the hardware components to server startups. And a pony. I want a pony.
At this rate by the end of the year systemd will be the only piece of software I will need to install on a machine. Since it seems to want to replace the entire operating system.
I'm trying to be ok with systemd but If I wanted to replace my entire operating system, and everything I know about that operating system. I would move back to OpenBSD. Atleast I know what I'm getting into there and how everything works is well documented.
We need to create a debian fork without any pottering taint. I mean jesus this is my career. Don't turn linux in win95.
http://boycottsystemd.org/ has generated quite a bit of traction for UselessD, a fork that tries to put the brakes on this flaming deathcab, but its interesting to see what the big 2 are doing. Ubuntu is committed to systemd because shuttleworth wants a "unified" experience and any users that care about their dwindling illusion of free will or choice have long since jumped ship. Pottering works at RedHat so theyve decided to hedge their bets that the server world, which is their bread and butter, is honestly interested in a binary logging monolithic pid0 that has udev and dbus as forced dependencies. user switching and networkmanager are fucking useless to me.
Leonard wont address this fact, but it stands and stands well. RC init was fine. SunRPC was fine. syslog was fine and consoles werent broken to begin with. Whining about mean developers in linux misses the point. you're redesigning something for the sake of redesign and its being done against the wishes of an ethos, the unix ethos, that has served well to maintain some of the most powerful computing systems in the world. There will be a pushback when you have the audacity to imply your linux redesign is a "choice" after a plurality of distributions adopt it.
Good people go to bed earlier.
Is systemd trying to become emacs?
There is nothing saying you have to be involved in the bickering. I own an iPhone, and I don't give two shits about Samsung vs. iPhone. You don't have to join a suicide cult to get religion.
> What is with the SystemD bashing anyway?
Like someone else already says. It violates the Unix design philosophy. It's a kitchen sink approach that creates unnecessary complexity, unnecessary dependencies, will be harder to work with, and will be more buggy.
If you want the Windows approach to systems design you can just use that. There's no need to pollute something else.
A Pirate and a Puritan look the same on a balance sheet.
Yeah, it was easier when I had to figure out what all this "daemon -x $DAEMON -p $PIDFILE" shit was, somehow translate that to the actual command being run, modify the init.d script, run it, try to catch an error, then go and put everything back the way it was and try again.
WTF is this "systemd journal $brokenservice"?
Support my political activism on Patreon.
... because it can never be early enough to buy some apps.
The last vestige of Linux has been removed from the GNU/systemd distributions, as systemd continues to move forward.
I propose new features to systemd to parallelize the hardware components to server startups.
Doesn't that already happen?
"First they came for the slanderers and i said nothing."
rabble rabble rabble!
Troll is not a replacement for I disagree.
... I don't understand why anybody cares...
Some people like to use software that is of a quality architecture and design, and not something that is little more than a security-challenged mash-up with very vocal protagonists.
The only answer I can come up with, as to why systemd is being shoe-horned into every mainstream Linux dist, is that Poettering is secretly employed by the NSA. And with Linux being what it is, moving forward, the NSA need a backdoor in place, since RSA and DUAL_EC_DRBG are suspect. If systemd becomes defacto standard for Linux, despite the criticisms that are falling on deaf ears of the decision makers, they'll have their 'in'.
The truly sad thing here, is that systemd isn't being deployed a path option. It's becoming, this way, or the highway (see move to *BSD). Which, for all my years supporting the FOSS and Linux environments, is wholly antithetical.
On the server side when admining hundreds or thousands of machines, troubleshooting a bloated needlessly complex system wastes precious time. Serious admins I know don't like systemd for the server side.
On the other hand, might be appropriate for desktop
Before you go round acting like the sky is falling, try educating yourselves about why this is necessary. This is not just a systemd problem, this is a problem for any init system that wants to support multi-seat, and sane switching between VTs:
Now you may say that OMG systemd is teh evil monolith!!1!!!, but before you do that understand that this has been an important feature that has been needed for a long time in any init system it just happened that SystemD solved it first.
Well, it's pretty simple. init had quirks - just like any other software. But they were well known.
I wouldn't care about the switch if everything worked in a systemd environment. Unfortunately, they haven't gotten all the possible software relationships figured out for what depends on what, and since it tries to start everything it can as fast as it thinks it can, system services don't start properly in some cases. It is hard to patch things so they work right since when the next systemd update comes out it rewrites the service policies, so if they didn't agree with your environment, you get to patch again. At times there have been enough changes under the hood that old patches don't work right. My approach was to move all the service start-ups to rc.local with appropriate sleeps or pings to attempt to ensure hardware a particular system relies on is running. Ugly and imperfect and probably prone to blow up in my face eventually, but my start up times are back down to a reliable 30 to 40 seconds instead of minutes - with the emphasis on reliable. Just reused the old init script numbering order to set up the service start order in rc.local and voila - it works. Can't tell you how many service network restarts systemd was trying on a couple of boxes before it finally could get started, but it was more than one. AARGH.
I'm sure in several years systemd will be as reliable as init was. Or they can add their own network stack and I/O handlers after they get the latest useless console functionality working and jettison the kernel. But the real gripe I have is they're worrying about things like the console before getting what they have working as rock solid as init was. Same gripe with Network Manager. It's great for a laptop with wireless or wired only. But the trunk and VLAN handling was an afterthought. It's gotten better, but that all should have been working BEFORE shipping the first test release to the users.
Article: Old, crusty, and possibly bug ridden part of the kernel is being moved to userspace. This new work will increase both the security and the stability of Linux systems, while adding the possibility of internationalization support.....
Slashdot Comments: Finally some one is doing something about CONFIG_VT. People have been bitching about that for years!
Article: this new feature is part of systemd.
Slashdot Comments: NOOOO! Why is Lennart taking away my freedoms! I'm switching to BSD.
It has gotten pretty clear that a lot of the hatred for systemd has nothing to do with the technical merits. This is a fix that has been a long time coming. Yet, almost half the comments are just more systemd hate fest.
[Disclaimer: Yes I hate systemd and I proclaim that loudly. Everything below is my personal experience with systemd and why I hate it.]
If booting the machine up was all it did, then I probably wouldn't care. Most of my hate (I can't speak for the rest of the internet) comes from the fact that systemd does a lot more. It also tracks user logins using a mechanism (control groups) that isn't available in some container scenarios making systemd unusable in those environments (and by extension any distro mandating systemd). It does its own logging in binary which needs a tool to read the logs and if it gets corrupted then systemd's devs say "just delete the logs". Really?
But I think the best reason people hate it is because it makes other applications become dependent on it. GNOME is the most well known example but I've also seen that Centos7's Source RPMs have systemd-specific commands (macros?) making it hard to build them on other platforms. rsyslog doesn't listen on /dev/log because systemd is doing something with the socket now. You cannot start services without systemd being the one to do it.
This is the hate. systemd isn't just an init system, it becomes part of your daily life. I liken it to the MCP (Master Contrl Progam) from the first TRON movie. It's systemd's way or the highway.
What is with the SystemD bashing anyway? It's written entirely for the Linux API for the better, why should Linux stick with the old init which runs in Unix System V and BSD. Isn't this why Linux is moving away from the old xorg to wayland.
Because especially in programming, "old" does not necessarily equal "broken." "If it ain't broke, don't fix it." And especially don't fix it with something that breaks half the rest of the product as a whole...while kind of failing to actually provide any benefits.
It will take some time to fully optimize SystemD for Linux and it will be the same thing with Wayland.
Except Wayland is still optional. You don't ram the new software into the working ecosystem and then fix it.
Unity? Screw that: XFCE. Slashdot Beta? Screw that: SoylentNews. Australis? Screw that: Pale Moon. UX developers DIAF
and did they move when canonical introduced "upstart" (you know, the ubuntu replacement for /sbin/init )? and did they move when canonical started to move to "systemd" from "upstart"?
"The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
I turn on the machine, it boots up, stuff runs, web pages get served an files get shared. Who cares what's handling the launching of startup processes?
You described that you are a user, those who cares are the developers and administrators.
Monolithic design makes it hard for developers to add functionality to without having an impact on those who don't use it.
Monolithic design makes it hard for administrators to tweak the system to be optimal for their needs.
There are always users that don't fit the majority, who wants to do something else than browsing the web and sharing files. They are the ones who gets screwed over by this because the answer they will get from the developers and administrators will be 'sorry, no can do.'
Read a sampling of the comments on literally any SystemD-related article to find out the details. It blows my mind that people are still posting "lol i have no idea u open sourc guys r dum" when all they have to do is scroll up or down like one screenlength.
Unity? Screw that: XFCE. Slashdot Beta? Screw that: SoylentNews. Australis? Screw that: Pale Moon. UX developers DIAF
"security-challenged mash-up" like running huge pile of barely maintained code in ring0?
You talk out of your ass, admining init on those is solved problem. You are the one with no knowledge and no experience.
The bloat of systemd impedes troubleshooting, and its bad design relying on far too many moving parts and binary logging database not readable by common tools makes it a project perhaps suitable for non-tech user's desktop (not embedded, not server) only
A few kernel developers are quite vocal in their hatred for it, by the way. The young kids with no large server experience might like it for their laptop
Yes, what every server admin wants is a situation where when they do a reboot for a critical server, they have to explain to their boss why the entire company's network is "down" for 10-15 minutes, and why half a dozen services didn't come up because they were using rc.d scripts that turned out to have shell scripting errors that the latest version of bash broke.
You are not alone. This is not normal. None of this is normal.
"Sys-V is an utter shambles"
Yeah, I wouldn't say I'm up on the SystemD controversy, but I would've thought that was the whole point. The old-style of system startup fires off so many processes, I've often thought you might be able to speed up a laptop boot quite a bit if you just wrote one perl script that did all of that stuff. It could even use the existing rc.d mess as an input format, read it all in once, and stash the info in a sane, single-file format...
-- Edsger Dijkstra
When dependencies become serious enough we end up with disasters like the poorly managed libc5->glibc2 upgrade. I fully expect the first update of systemd will absolutely take down systems when the upgrade tries to restart some service everything depends on, including the upgrade process.
SystemD fires off more processes, and moreover parallel starts services (which can be good thing, but that can be done with init style sytem too, long solved problem)
year of GNU/Hurd on the desktop!
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
The new broken way of doing things means that my more interesting Ubuntu machines don't boot up without user intervention despite there being no actual problem.
Attempts to fix this myself will be more difficult because the newer system is more complex and thus more prone to user error.
Simple and reliable is not a bug, it's a feature.
A Pirate and a Puritan look the same on a balance sheet.
Well I get shiny RHEL7. I build my systems like RHEL6 (I don't do a lot of custom init stuff).
RHEL7 fails to boot after a while because of systemd. No one knows why. At first I was told it was hardmounts. Take them out in rescue mode, no booty.
Systemd by design tries to mount nfs shares, before it even starts up the network, out of the box! Systemd supresses everything unless you tell it to. Why? Because some hotshot idiot thought I was using RHEL to run a desktop? Oh hey, just what I wanted BINARY LOGS THAT BREAK ALL MY EXISTING AUTOMATION.
This is the problem with systemd, it is unportable, monolithic and subject to dictatorship. And what does it offer me day to day in a grinding development lab? Nothing. Also yes, Embedded systems. I still support those too.
---Up Up Down Down Left Right Left Right B A START
so what?
Shifting things from kernel space to user space is usually a good idea, especially with respect to security. Of course the ttys should not stay in PID 0 but be moved to separate, user specific processes in the long term. But moving them from kernel space to PID 0 is an important first step in that direction.
>Sys-V is an utter shambles when automating that many machines.
That's funny it works for my broad spectrum workloads and systemd does not. Why do I not have a choice?
---Up Up Down Down Left Right Left Right B A START
What is starting processes isn't so much the issue. The issue is that systemd is demanding major changes of other software to work with it, and this is then making this software non-portable. e.g., Gnome 3.
People don't want to run Linux everywhere. Despite what some people think it is not always the best fit. There are other Unix platforms which fit better. Platforms that have had application compatibility with Linux up to this point. systemd has changed this. Changes to Gnome to work with systemd for example have made it non-portable to other platforms.
This is a problem for anyone who wants to say, develop a cross platform gnome3 application.
That, and there are the corruptable binary logs, the solution to which in the bug report is to "just delete them" and the bug has been closed as won't fix. Sorry, but if this is the resposne to journal corruption rather than finding out WHY the journals get corrupted and fixing the fuckign problem, then i do not want that in control of my logfiles.
Also, the massive violation of the KISS principle that has been a core guiding principle of Unix design since the start.
Systemd is a poor solution to a non-problem. There are plenty of other problems to tackle first, before trashing and re-writing working, well debugged code and breaking cross platform compatibility for no good reason.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
FreeBSD is my choice of BSD, it's a little more versatile and better supported in my opinion.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
Well he started it!
("MOM! He's on my side again!")
Seriously though - I can't help but think the only real difference between here and, say, Microsoft is that you can actually SEE it here instead of it being locked up behind the office doors.
Hacker Public Radio is our Friend
No. I want it to go back in time a couple minutes before grub even runs to magically make all the bios stuff happen in parallel. I will accept nothing less. And I want binary consoles instead of just log files so that I can feel like I'm in the matrix.
And I want binary consoles instead of just log files so that I can feel like I'm in the matrix.
Well that might actually be really cool
"First they came for the slanderers and i said nothing."
Really? Where were you when microsoft forbed Vista off on everyone? The IT guys where I work spit flame and venom every single time someone uttered the term Vista. I learned a lot of new vocabulary during that period. I've spent most of my life in or working for the military and I was impressed by the obscenity and profanity laced invective.
There is some virtue in "tried and true." I think systemD may not be as bad as a lot of people make out but then I don't really know that it was all that needed either. Either way though I can't tell much as a user that it has changed under the hood of the latest linux distro I run. Currently running SparkyLinux which is a Debian derivitive.
You do realize that this whole stalking and multiple-post tactic only makes you look pathetic and kooky and does nothing to discredit your target du jour, right? This isn't how debates that people take seriously are carried out.
I doubt the AC you responded to was BarbaraHudson. Most of us here are sick of seeing tons of identical off topic replies from you.
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
Are Linux using companies rushing to get RHEL 7? Are Linux users anxious to upgrade to Systemd based systems? Or, has Linux with Systemd proved unpopular?
Me too.
Remember when we moved from a.out to elf executable format? That was a lot of work but we all could see the benefit and although software would break, everyone would know why.
I think I'm with you on this. Compared to Solaris, HPUX and even Aix, Linux has no place in an enterprise looking to keep costs down.
Apart from instances like now when I'm running four X sessions at once because it makes some full screen software behave a lot better and sandboxes workflows nicely. Then there's the people that were running multiple keyboards, mice and monitors for a couple of people on the same box well over a decade back. What is it with Wayland fanboys not having a clue about the system that they are supposed to be improving on?
In addition to the other reasons people have given you (bloated, breaks the unix idea of doing one simple thing well, binary log files, doesn't play nice with others, etc), the reason people are rabidly opposed is because of the way it's being adopted, or should I say, thrust upon us. Poettering and friends are not simply making a piece of software and releasing it and getting people to adopt it because it's good and solves a useful problem. They're playing shady political games to force adoption.
Ideally, if you think you have this great new replacement for the fundamental piece of userland software in Linux, awesome! Write it, fork a distro and build your distro around systemd. Use it. Find the bugs. Work them out. Do this long before you start suggesting people run it on their servers. If it's actually better, distros will start including it. Instead they've played political games to force it into really popular distros like Debian.
It's just not ready for prime time. Build it, test it, show us how it's better and we'll be overjoyed to use it. But ram buggy bloated bullshit down our throats for no other reason than your own ego and well, fuck you.
We don't have a state-run media we have a media-run state.
Some people like to use software that is of a quality architecture and design, and not something that is little more than a security-challenged mash-up with very vocal protagonists.
Right. But those people abandoned Linux for OSX years ago.
The only time I see people round here unable to use capital letters to make sentences, they are posting Linux positive stuff.
I'm done with Linux. Screwing around too much with stuff that doesn't need to be messed with is giving me headaches and sucking up more of my time that I can better spend on other pursuits.
Welcome to OSX. The OS people go to when they realise constantly fixing and customising computers isn't the point of having them.
unfortunately these trolls don't have the mental capacity to realise they are behaving like twats
"The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
It violates the Unix design philosophy. It's a kitchen sink approach that creates unnecessary complexity, unnecessary dependencies, will be harder to work with, and will be more buggy.
Sounds like EMACS. Who knew that violated the Unix design philosophy. Has anyone told RMS?
No, all the extra firmware will load one device at a time. So you get to see all 4 NICs load their firmware, your onboard SAS controller loads its RAID stuff, then your fibre-channel HBAs, your BIOS does its thing, and so forth and so on, not necessarily in that order, and all in serial.
One of the reasons things are done this way is that generally each device's firmware will give the operator an opportunity to modify settings as it boots, so it will ask you to hit a key or some combination of keys and then wait a number of seconds for that to happen.
If we had some fancy modifications to LinBIOS maybe you could just tell the BIOS that you want to get into the firmware options for X device and then it would take you right to it, while skipping all the waiting for input.
But, alas, we don't, so we just wait for 5 or 10 minutes while the server boots.
The worst part is when shortcut keys and mnemonics are changed because of a different language, I navigate by keyboard and have to re-learn everything on other systems and guess what the right command is if I don't know the language.
Twinstiq, game news
systemd is optional, you can remove it and replace it with any init system you like
"The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
It is pretty sad to see, that after so many comments nobody really has a clue about what the story is about, and what is happening in the Linux kernel.
The kernel VT system has been considered a monstrosity by kernel developers the last decade and everyone is of the opinion that it should be used to user space.
The finally a really smart guy actually attacks and solve the problem. His name is David Herrmann, and he has tirelessly worked on this for years. Systemd distros will get the full support of his research, simply because almost all Linux distros are using, or a going to use systemd. But don't worry, he has provided rich support user space VT's on non-systemd Linux distros, by eg. "ksmcon"
https://github.com/dvdhrm/kmsc...
Here is his fosdem talk:
https://archive.fosdem.org/201...
Here is his blog that will tell you more about VT's than you ever knew:
http://dvdhrm.wordpress.com/
Here is a wiki link about VT:
http://en.wikipedia.org/wiki/L...
Here is an old blog post about the problems with the old kernel VT:
http://dvdhrm.wordpress.com/20...
In short, no need for the systemd opponents to get their panties in a bunch; they can either use Hermanns user space tools, or pretend there isn't a problem and use the present kernel system.
For the rest of us who really likes systemd, this is great news. Thanks to Hermann's work, there will be much better console support for early boot debugging, better security, better keyboard and language handling etc.
Yeah, and then GNOME (and anything else that depends on it) won't work. Try to keep up.
Unity? Screw that: XFCE. Slashdot Beta? Screw that: SoylentNews. Australis? Screw that: Pale Moon. UX developers DIAF
I'm checking it out too. At this point I'm mostly concerned about servers though (and simple servers at that, I don't need a lot of third party software) so I think OpenBSD is a good fit there. That's what I was using Debian for all these years because of it's stability. That said even before systemd came along Debian was actually starting to bug me slightly with changes between distros.
On a desktop I don't care how fast a distro changes, I like mucking around with my desktop. And whether it was throwing Mint on because I knew it would, for the most part, Just Work or throwing Debian on a laptop and bothering with hand crafting a desktop I never worried about what was under the hood too much, just enjoyed seeing GNU/Linux mature as an end user solution.
But I don't want any of that on a production server. I want something known for stability and security and something that a mere mortal like me can understand with regards to the base system. OpenBSD fits this criteria nicely.
Despite my previous comment it's not like I'll never use Debian or GNU/Linux again, I'm just replacing my server infrastructure with something I can depend on staying consistent and stable for the foreseeable future.
Thanks for the advice!
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
It violates the Unix design philosophy
If the Unix design philosophy is to have cryptic init files scattered all over creation then it needs to be violated.
I don't get all the hate for systemd, because I don't understand why anybody cares. I turn on the machine, it boots up, stuff runs, web pages get served an files get shared. Who cares what's handling the launching of startup processes?
Because some of us worry about when things don't Just Work. Some of us have jobs whose responsibility it is to Make It Work. And systemD makes that more difficult at times. For example: their binary logging system ("the Journal") is not ACID compliant and can become corrupted if there's a (unrelated-to-systemD) crash of the system.
Now if the system is crashed, I want to know /why/. But if the logging has been trashed, how can I get any diagnostics of the last few moments before the system went AWOL?
The systemD developer response to corrupt logs? "NOTABUG":
https://www.libreoffice.org/bugzilla/show_bug.cgi?id=64116
So instead of using a decent back-end data store for logging (e.g., SQLite, BerkleyDB, OpenLDAP's LMDB) they decided to reinvent the wheel and did a piss-poor job of it. Then they deny it's a problem.
https://www.linkedin.com/today/post/article/20140924071300-170035-why-you-cannot-trust-lennart-poettering-systemd
I have not (yet) used systemD, so don't know if it is the greatest thing since sliced bread or a complete load of crap. It purports to solve a lot of problems, but I've never experienced them and so the change to me seems unneeded. Add the above problem about logging, and I don't have a lot of faith in the design of the thing.
I haven't seen a lot of pressure yet. With RHEL6 the urge to switch came on fast. Esp with regards to virtualization and what not. I haven't gotten the sense that RHEL7 has a whole lot of "must have now" tech in it, as opposed to the amount of systemd fear it has. Based on some quick googling, systemd can be configured to send your log files over syslog so that takes care of most of my corporate compliance concerns. I don't think the systemd folks have done a good jobs educating people on how it can fit into existing workflows. Esp. against all of the noise out there. I don't really know if it's even possible to fit it in with everything. But I guess we'll find out. I currently have no opinion, I see some benefits as well as some drawbacks.
This is real. This is the result of more than a year of work. Blog posts have detailed all of that for months if you knew where to look (I didn't, but friends of mine pointed me to this). While I can see the need to clean up the existing interface, I'm not a fan of the resulting solution, either.
...>systemd brings order and consistency. None of the kernel devs are bothered about the change, and most already use it. Whether you like it or not, it is now the de-facto management for all of the major distros for new installs.
Don't like it? No one cares what you think.
Is that you, Lennart?
'The Economy' is a giant Ponzi scheme whose most pitiable suckers are the youngest among us and the yet-unborn.
systemd is optional, you can remove it and replace it with any init system you like
Unless you are using a disto like Gentoo, you can't replace systemd because everything useful depends on it either directly or indirectly.
And now, I expect that you will say "that's not systemd's fault, that's the fault of the distro". Except that without dbus, systemd won't run, and without dbus, you the number of apps that will actually run can be counted on on hand. And, it's that way because the same guy that wrote systemd wrote dbus, and baked in the dependency.
OTOH, if you have a step-by-step guide on how to replace systemd with any other init system on current releases of CentOS or Fedora, and the upcoming release of Debian, I'm sure there would be many people who would be grateful.
If you care about laptop boot performance, get an SSD. Mine boots in under ten seconds with the awful old init files. So systemd might shave a couple of seconds off that. Who cares?
And boot time is irrelevant for servers that spend five minutes in the BIOS before they even think about loading the OS.
"It does its own logging in binary which needs a tool to read the logs and if it gets corrupted then systemd's devs say "just delete the logs". Really?"
Er...no. They don't say that. journalctl reads as much data as it can from corrupted log files and otherwise routes around them. I don't know of any advice that says to delete them.
journald is also intentionaly designed to make it simple to store logs in plain text format if desired, using rsyslog or something similar as a journal consumer. you can do this alongside or instead of systemd's native log format.
um. what? people have been updating sytems that use systemd for, what, four years now? amazingly enough, it actually works, we don't just all watch our systems crash and reboot every time.
This approach has it's merits but it's become clear that it's a thing ill suited for many of the challenges linux faces in 2014.
Like what?
The current init system is a clunky mess, but it works because it's been kludged up over decades. What single benefit does systemd actually bring me?
"Systemd by design tries to mount nfs shares, before it even starts up the network, out of the box!"
Um. No it doesn't. Who told you that?
"Systemd supresses everything unless you tell it to."
No it doesn't. Who told you that?
Sys-V is an utter shambles when automating that many machines. You obviously have not experience of it if you claim otherwise.
We don't have "thousands" of machines yet where I work, but are over 1,000. With the rise of virtualizaton, SysV does just fine, since most servers are now single purpose, and don't have a lot of interdependent services running. For these, systemd really is a solution in search of a problem.
OTOH, for a GUI desktop which starts "services" when a user logs in, then systemd might help a lot.
"Oh hey, just what I wanted BINARY LOGS THAT BREAK ALL MY EXISTING AUTOMATION."
systemd is designed to make it trivially simple to have text logs if you want that. RHEL 7 is configured by default to do permanent logging in plain text format via rsyslog; the native journald logs aren't even permanently stored by default (this is the config that was in Fedora for a while before journald's native format became the default/primary).
https://access.redhat.com/docu...
I am starting to suspect you're a troll and haven't actually used RHEL 7 at all.
"The only things I use Linux for anymore is for my personal file server in my basement, and even that is running old Fedora 17."
so...it's running a rather early version of systemd, and apparently working fine, because you haven't seen a need to upgrade it?
Yup. That's why the major distros are all switching away from SysV init to systemd. Because the alternative is to continue with a Rube Goldbergish mess of shell scripts with no proper event-based activation nor decent process monitoring.
I am most definitely not the only one fed up with chasing Heisenbugs in a maze of twisty little shell scripts, all alike.
"I know I will be modded down for this": where's the option '-1, Asking for it'?
https://access.redhat.com/docu... is a rather comprehensive and good guide to the changes in RHEL 7 for RHEL sysadmins. One thing you'll learn from it is that RHEL 7 logs to text via rsyslog (as a journald consumer) *by default*.
Thirty seconds every six months on a system where the motherboard BIOS POST, each NIC firmware, the SCSI card firmware, spinning up the drives, and the RAID bios take around two to three minutes to complete.
You mean thirty seconds on a container that otherwise starts up in half a millisecond? You do realize that nobody runs servers on bare metal these days, right?
I haven't seen a lot of pressure yet. With RHEL6 the urge to switch came on fast. Esp with regards to virtualization and what not. I haven't gotten the sense that RHEL7 has a whole lot of "must have now" tech in it, as opposed to the amount of systemd fear it has.
We're a mixed shop, running CentOS 6 and Ubuntu LTS (12 and upgrading to 14). We will not be upgrading to any version of either distro that has systemd until at least a year after release, and maybe even later, despite the fact that upgrades to LTS 14 happened within a month of release for many of our systems.
Thanks. I've read this before. But not currently switching to RHEL7, so have obviously forgotten everything that isn't relevant to my day to day trials. But thanks again for the reminder.
To be fair, upstart doesn't have anywhere near the level of "kitchen sink" insanity that is systemd.
God is imaginary
You are funny, systemd is a beowolf cluster of Rube Goldberg contraptions with fractal self-symmetry in its hieracharcy.
For any other Free software, the default response to complaints is "So where are your patches, asshole?"
You do realize that nobody runs servers on bare metal these days, right?
I'll grant you that print servers and web servers tend to be on VMs these days, but the VM host OS has to run on bare metal, and if you're doing number crunching or large data storage, it's going to be bare metal too, because the VM host/guest overhead is a waste of resources.
No, as you point out yourself, systemd is moving toward Windows, while the rest of the world is moving away from it.
simple, just install the things that don't depend on it
"The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
If the developers of "everything useful that depends on it" have updated their "xxx" to interact with it, just appeal to the developers of "xxx" to make the dependency optional.
i have no interest in replacing it as it works for me but opensuse does have a sysvinit package you can install. I'm relying on the very vocal detractors of systemd to be sufficiently competent to install the init system of their choice because they are adopting the position of saying they are smarter and better than the developers of systemd.
"The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
In Russia, Unix design philosophy violates you!
As far as I can understand they are fscking serious.... OK guys, so long and thanks for all the fish. Any variant of BSD (OK, except the shit Apple is selling) for my cup of tea...
systemd managed to replace init, inetd, and some of cron in what appears to be a stable environment. This allowed systemd to work in docker and drastically improve Linux virtualization to leapfrog Solaris zones.
What systemd did not do was provide reasonable documentation. RedHat's v7 inittab has a website for a blog post that sucks. There is no general intro for users attempting to create crontabs executed by systemd, inetd entries for common services, and runlevels that control groups of processes.
systemd fell down hard on documentation, and the first blush with the unix admin crowd has not been kind.
These developers delivered working code in a radically new environment, but without documentation the architecture appears to discriminate against people who have been doing things the same way for 30 years. The authors, and their software, appeared cliquish and discriminatory. Had the software and the documentation enabled a gradual migration into a more powerful architecture, things would have been quite different.
In any case, this is no justification for people to be vile. The old crowd needs help into the new environment. This help needs to happen, and the insults and threats need to stop. Both sides need to work together to get us where we need to be.
I think I'm glad to read this news, especially right now, because it might motivate someone to develop a better alternative.
I'd like to see something like this:
I do embedded development, and most of those have no GUI or X server. I can use VTs to keep VIm sessions open on a few (yes, I know VIm supports multiple files/buffers, but VT switching is easier), a VT for compilation/test, and, with GPM, copy/paste between them.
> using rc.d scripts that turned out to have shell scripting errors that the latest version of bash broke.
Those scripts shouldn't have been passing functions by environment variable in the first place.
I should use this sig to advertise my book ISBN-13 : 978-1501515132.
Doesn't have a damned thing to do with Windows or binary files, it has to do with the fact that Debian has been made Red hat's bitch by way of ex RH and Ubuntu employees taking over the board. For those that want to know what systemd is REALLY about its about cloud computing, specifically RH is pushing cloud computing like mad and systemd is gonna end up being a "SVCHost" for Linux dedicated to managing cloud computers.
This is one time me and the FOSSies are actually on the same page, as just like windows 8 was forced from on high and gave the users a big fat greasy finger so too is systemd being pushed by corporate with exactly zero fucks given about what the end users want. Ironically despite all this "empower the user" talk Linux has always had this is one case where Windows users had more power thanks to the ability to vote with their dollars, thus getting Win 8 shitcanned in favor of a much saner and nicer Win 10. But this does not mean that all hope is lost in Linux land, it just means you are gonna have to organize and SCREAM BLOODY MURDER and refuse to take this bullshit. You especially have to organize all the volunteer coders and get them to walk away, because losing all that free labor and forcing Red Hat and friends to pay for every single dime's worth of work is the ONLY way most of you can hit 'em in the pocketbook. those of you that run non cloud based servers can of course tell them you will no longer use their products but considering how much time and money you have invested in your servers I really don't see that happening.
Finally you need a rally cry, something simple and catchy and on message to focus the narrative and rally the troops, a "fuck beta" for systemd if you will. And since old Hairy will ALWAYS stand for the users allow me to give you one as a show of solidarity in your plight. Its simple, concise, on message, and sums up in a single simple sentence WTF is wrong with systemd..
SYSTEMD...Its the Metro of Linux!
ACs don't waste your time replying, your posts are never seen by me.
Sounds like a Feature, not a bug... Replace that Gnome with something worthwhile.
"But I think the best reason people hate it is because it makes other applications become dependent on it."
No it doesn't. Gnome didn't have to use systemd at all, it was a choice by the Gnome team.
Considering that it's posted under AC I'd say that's affirmative.
Yeah, I'm an XFCE guy myself ever since Unity, but anything that depends on GNOME (e.g. GIMP) may be problematic.
Unity? Screw that: XFCE. Slashdot Beta? Screw that: SoylentNews. Australis? Screw that: Pale Moon. UX developers DIAF
It's not Metro.
It's SVCHOST.EXE
But that's a little too arcane, for people who neither debug their own system, or who are not security specialists of one stripe or another.
How about "systemd is a BSOD" :-)
https://lh3.googleusercontent.com/-bZId5j2jREQ/U-vlysklvCI/AAAAAAAACrA/B4JggkVJi38/w426-h284/bd0fb252416206158627fb0b1bff9b4779dca13f.gif
"Flyin' in just a sweet place,
Never been known to fail..."
One of the defects constantly levelled against systemd is its propensity to corrupt its own system logs, and how the official response to this defect is to ignore it. The uselessd page has a link to the bug report in question, which was reported in May 2013 and, over a year later closed and marked NOTABUG. However, it seems Mr. Poettering is getting annoyed by people using his own bug reports against him, and added a comment to the bug report today purporting to clarify his position.
Unfortunately, his "clarifications" serve only to reinforce my suspicion that systemd is a thing to be avoided. To wit:
Well, yeah, corrupt logs would be regarded by many as a major bug...
Okay, so freeze the corrupted data set so things don't get worse, and start a new data set. A reasonable defensive practice. You still haven't addressed how the corruption happened, or how to fix it.
Okay, so journalctl tries to be robust, assumes the journal data might be crap, and works around it. So we can assume journalctl is probably pretty solid and won't make things worse.
....Uhhhhh-huh. So, yeah, newer tools will do a better job of working around the corruption, and we'll be able to recover more data, assuming we kept known-corrupt logs around. But what I still don't understand is WHY THE LOGS ARE CORRUPT. And why aren't there log diagnostic and analysis tools? If you already know your logs can turn to crap, surely there are structure analysis tools around that let you pick through the debris and recover data that your automated heuristics can't.
And why do I get the feeling that implied in the above is, "You don't need to know the log structure or how to repair it. We'll write the tools for that. We'll release better tools when we get around to it?"
....AAAAnd you lost me. Seriously, this is your defense: "Filesystems are more important than system logs, so they have to try harder?" I find this insinuation... surpr
Editor, A1-AAA AmeriCaptions
as the operating system of choice.
Seriously what happened here?! Yes those corporations that are in the business of making profit will do what is in their best interests rather than the users. We know that, we've always known that, so what we have always done is to take the systems they provide - that we want/need to use - and work around any issues that we don't like. What they have provided out-of-the-box has never been ideal for everybody so the geek/nerd way of dealing with that (and then helping others deal with that) was to come up with new ways to work while still exploiting these systems.
With Windows we've had awesome alternative shells like LiteStep, Emerge, bbLean, etc... And when Open Source projects start to be driven in a different direction they are forked (MySQL/MariaDB for example or Open/Libre Office) or you switch to an alternative that does things the way you like (for example Slackware and Gentoo don't depend on systemd). Yet at some point it all devolved to stories about Windows 8 or Gnome3 or Ubuntu or systemd being just hundreds of angry comments from detractors venting rather than genuine solutions - which do indeed exist - to the problem. Nowadays it's mostly just people getting angry and throwing their hands up in the air shouting "it's hopeless". The Amazon/Ubuntu issue was a very simple fix and if you took issue with it such that turning it off wasn't enough then it would be trivial to maintain a fork with that code removed or a patch to remove it but still those who claimed to care about the issue were more interested in posting "fuck you canonical" than fixing the issue.
It's not all hopeless but it's not all going to be done for you either, getting all angry about it is just detrimental to you. Coming up with a solution to a problem you have is rewarding and sharing that solution should be even more so but the response around here is most often "well I shouldn't have to do anything else, it should work out of the box" well wake the fuck up, corporations don't act in your best interest and never have and never will.
It's a Lennart Poettering meltdown and accurately exemplifies the problems people have with him, those around him and why people are justifiably concerned about the software he is being allowed to write. This guy simply has 'issues'.
If he believes he's going to take on Linus and the kernel developers, stamp his feet and take over the Linux landscape he's going to lose badly.
Wake me up in few years. It's still too early to adopt SystemD.
I mean, the damn stuff looks like it's in alpha, what with lacking basic stuff like it's own filesystems and network drivers. :) :D
When it gets to the point where it can update the CPU's microcode, I may look into it.
You do realize that nobody runs servers on bare metal these days, right?
I'll grant you that print servers and web servers tend to be on VMs these days, but the VM host OS has to run on bare metal, and if you're doing number crunching or large data storage, it's going to be bare metal too, because the VM host/guest overhead is a waste of resources.
Then you'll be happy that systemd ensures that your host and container service managers and loggers are well-integrated. :)
And containers have very little overhead. They are just processes.
And how often are you going to reap that reward? I just did a quick spot-check of our 1,200+ host local VM farm, and didn't see any hosts with less than 60 days uptime; it would be longer if not for quarterly patch/reboot cycles. You realize that's sort of the entire point, right?
When I run updates on a container I shut it down, snapshot its subvolume, and start it up. That takes a few hundred millseconds with zero impact to anything that depends on it (nothing times out that fast). Why not do it - the reboot costs nothing and gains you a bit more consistency if you have to roll back.
Your sysadmin practices are based on the assumption that reboots result in downtime.
SUCK IT RED HAT & DEBIAN LOSERS! (..and derivative losers!)
Your thin skin doesn't make me a troll
Am I alone in wondering who would benefit from chaos and strife in the Linux community?
As always, all IMO. Insert "I think" everywhere grammatically possible.
systemd would make a great OS if it only had a decent init system
https://bugs.freedesktop.org/s...
Indeed, I misremembered that. They don't say delete, they say the file gets rotated out immediately. And this bug report is famously linked as a demonstration of why systemd is hated for its developer attitude to the point that Lennart repsonded to it (today oddly enough). Corrupted files are not considered a bug and not getting fixed.
If he believes he's going to take on Linus and the kernel developers, stamp his feet and take over the Linux landscape he's going to lose badly.
I am hoping for that to happen. It is basically the only thing that can still stop this madness at this time.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Other than being forced to type in 12 passphrases manually to decrypt each hard disk at every single goddamn boot, because custom keyscripts just "aren't the systemd way". Or spending hours figuring out why your Requires=network.target units inexplicably never start on boot, without a single shred of clue or evidence or event in any logs whatsoever, despite LogLevel=Debug and even though network.target clearly flashes by during boot and systemd-analyze clearly shows that it knows about this relationship with your unit and the service starts normally when you login and systemctl start manually. Or that tweaking your daemon args now requires a systemd daemon-reload as well as restart.
Yes, apart from all that, and the time saved now that admins will never have to see another freaky, alien shell script ever again because init systems were the only thing which used them, apart from all that... I'm hoping like hell systemd will hopefully one day buy me something other than more downtime.
Oh, how embarassing.... the above post wasn't intended for you.
Because just like in politics Red Hat figured the best way to ram their shit through was to own BOTH choices?
Like it or not the little distros like Slax just don't have the resources to go it alone, which is why damned near every distro on the planet is based on Debian or Red Hat. Now thanks to Red Hat loading the Debian board with ex RH and Ubuntu employees whatever Red Hat wants to push WILL be in Debian, like it or suck it users! And as we have seen this isn't gonna be some minor subsystem one can fork out, with each release they are gonna tie more and more critical subsystems into it so just like SVCHost in Windows there is gonna be no damned way to get rid of it without crippling the OS so bad it will take millions in cash to build up a large enough dev team to rip that shit out AND maintain all the no longer supported software that was there previously AND be able to reintegrate that software AND be able to deal with the new pieces that are absorbed by systemd...again NOT for the benefit of the community but ONLY for the benefit of Red Hat's cloud computing platform.
So I'm sorry but the "just fork it" bullshit ain't gonna work because you have no base to fork it FROM, it'd be as impossible as saying you don't like the way NASA is headed so you are gonna take this raw ore straight from the mine and create a Saturn V out of it! You might as well try to build a new multitasking OS from scratch, I seriously doubt it'd be any more difficult than ripping out the entire Linux lower stacks and rebuilding them from scratch with unsupported code!
At the end of the day unless you can get the entire community to do something radical, like stop developing for Linux completely and throw their weight behind BSD or another alternative? Then you are gonna have to face facts, RH has fucked Linux in the ass and its only gonna get worse. As one dev put it "Linux is quickly becoming what I left Windows to get away from" yet sadly at least if enough Windows users vote with their wallets they CAN affect some change, RH has made it clear they do not give a single fuck about what the community wants, this is about selling their cloud, period.
ACs don't waste your time replying, your posts are never seen by me.
Like it or not the little distros like Slax just don't have the resources to go it alone
Slackware plenty of support and it doesn't use systemd. But that is what the community is for, if enough people aren't willing to contribute time or money then clearly the project is not viable.
whatever Red Hat wants to push WILL be in Debian, like it or suck it users!
So desktop Linux has only ever existed at the whim of Red Hat?
RH has made it clear they do not give a single fuck about what the community wants, this is about selling their cloud, period.
Oh wow you mean a for-profit corporation cares more about its shareholders than the "community"?
What is with this "I want Linux and I want it my way and I want Red Hat to do it for me" attitude? What is it you think they owe you?
Yeah I going to really miss my bash script based terminal... these new binary terminals are really terrible. /sarcasm
Perhaps they should just drop the 'd' in the name and write their own kernel to go along with their... thing. Voila, no more problems with ill-tempered Linux kernel developers! And they could all integrate it in one huge, funky ball o' bits!
If a train station is a place where a train stops, what's a workstation?
Now that's a totally practical suggestion! I'll get right on it!
If a train station is a place where a train stops, what's a workstation?
Never mind using the word, cabal. I'm sure Mr. Poettering is very aware of its scary overtones and, in a mis-placed sense of mischief, he's push everyone's buttons with it.
But have you read the proposal there? It looks like "the cabal" means to have systemd as a virtual-box-hypervisor-like entity, able to select OS and version per login. Am I mis-reading something? This looks like 2nd System Effect gone absolutely berserk! I'll have to re-read the article to make sure his occasionally crusty English hasn't kept me from understanding something positive that "most users" really want whether they know it yet or not.
cheers...ank
Still hoping for Gentle Treatment...
whatever Red Hat wants to push WILL be in Debian, like it or suck it users!
Note WILL, in capital letters. Typically used to denote future tense.
So desktop Linux has only ever existed at the whim of Red Hat?
Emphasis mine, on "has" and "existed". Typically used for past tense.
Take a reading comprehension class first.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
That, and there are the corruptable binary logs, the solution to which in the bug report is to "just delete them" and the bug has been closed as won't fix. Sorry, but if this is the resposne to journal corruption rather than finding out WHY the journals get corrupted and fixing the fuckign problem, then i do not want that in control of my logfiles.
Log files are unimportant to the people who are writing and advocating for SystemD.
Logs files are criminally important to those who have to report to regulatory agencies and otherwise important to those who manage systems for others (businesses and governments). None of those agencies will accept missing, incomplete, or corrupted log files during an investigation.
The people writing this software need to get serious or get out. Logs are of vital importance.
It seems like I am surrounded by inexperienced children who have no adults to discipline them.
"Someone needs to talk to the tree of liberty about its ghoulish drinking problem." by ohnocitizen
that an OS architecture that dated from the 1970's was actually totally elite
Well, humans needed to breathe in 1970s, and they need to breathe now. So just because something is from the 1970s clearly cannot be held against that. In fact, we have 44 years (ignoring pre-1970 of course) of evidence that stopping to breathe is a bad idea, and 44 years of evidence about which parts of UNIX philosophy are applicable when and why.
Now your long post didn't address any single problem with the UNIX philosophy. Apple clearly showed that integration for desktop users is not impossible with UNIX, and UNIX philosophy is not even against a user exposed integrated interface.
I will be the first to admit that at times UNIX philosophy is not applicable - e.g. ZFS combining LVM, raid and filesystem in one single monolithic feature is against the UNIX philosophy. But it solves many real problems, without introducing new ones. And it wasn't considered stable for just under a decade after release.
Systemd was considered "stable" within an year or 2. The parts where it breaks UNIX philosophy are clearly where it is NOT good to break it, with a nice bug to show for it.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
So a bug in systemd now (or botched update) would be a totally unresolvable, because Linux box wouldn't even have proper console then?
Very very sane plan. Kudos, systemd. /s
All hope abandon ye who enter here.
like it or suck it users!
I have impression that only I recognize the trend: the Linux is slowly transforming into the UNIX.
The same crappy attitude. The same crappy unusable bundled software you can't unbundle or replace. Admins are overjoyed with the new shiny toys - users are making teeth screeching sounds.
All symptoms support the diagnosis: it's UNIX.
All hope abandon ye who enter here.
But I think the best reason people hate it is because it makes other applications become dependent on it
Not at all. Not even slightly. Gnome wanted to use control groups. One API provided by control groups is through logind. Logind is dependant only partially on systemd, and in some cases (Ubuntu) it's actually possible to install Gnome 3.8 and logind without installing systemd.
Systemd provided an API for cgroups through logind. It was the Gnome team which chose to use this API, and that's all that it really is, an API, . If people are so dead set against systemd then by all means ban together and write an alternate API for cgroups for Gnome to use. The only reason that Gnome has a dependency on systemd/logind is because the Gnome team has better things to do than reinvent the wheel.
And before someone blindly starts spouting out some nonesense about Gnome shouldn't be dependent on others, please check your history to make sure you're not the same people blindly spouting out "do one thing and do it well" as a retort to systemd, when really that philosophy is quite dead in a modern Linux system.
you can do this alongside or instead of systemd's native log format.
Not using syslog independently - but only through journald. Why such an abomination was considered a good idea, you know better than me.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
Incredibly nasty hacks to make an init system work in parallel is not "solving" the problem. SysV init does NOT support parallel starting of applications, and many of the bolt ons that have enabled this function do not support event based starting of apps either.
If your long solved problem wouldn't still be a problem there wouldn't be masses of teams working on many fronts (not just sys-v but upstart, daemontools, etc) to "fix" Sys-V init.
Please join my group.... :-)
https://www.facebook.com/group...
Actually no, this is informative to me. Thanks.
I am not a troll. I am frustrated because RHEL7 is a heaping pile of poop presently.
---Up Up Down Down Left Right Left Right B A START
Yes. I'm beginning to wonder if Adam uses RHEL7 outside of 'vanilla' norms, very frustrating although I'm sure the kinks are worked out.....but I'm still left with WHY did we break it? Couldn't we have made systemd an option not a requirement?
---Up Up Down Down Left Right Left Right B A START
Anyway my point is that the above poster made an utterly huge newbie mistake that showed he's a bit of a backyard dabbler instead of someone that has seen X in action in a workplace - he was pointing out a limitation that does not exist and saying that he can solve that non-problem. I'd like Wayland to be nice but these guys need to get out of their "X sux" rut and take a look at why people actually use the thing. There's been a dozen failed dumb framebuffer projects due to not making it work in such a way that developers were interested in writing or porting software for it, especially since it's a lot harder writing stuff for a supposedly very fast dumb framebuffer than something with a bit more complexity that handles a lot of output instead of the application having to do it.
I'll give you a quote from the often quoted on Wayland Daniel Stone - "Multiple screens - that's an application problem". Bit of a worry, however others in the project may have very different views, he wants it for phones after all, so it won't necessarily end up stuck that way. Wayland may expand to provide enough stuff to be useful despite some of those things being initially dismissed as "crufty X stuff".
For someone not actually involved in the development of either that's a pretty weird thing to write.
I'm relying on the very vocal detractors of systemd to be sufficiently competent to install the init system of their choice because they are adopting the position of saying they are smarter and better than the developers of systemd.
No, we're adopting the position of "if it ain't broke, don't fix it" and "provide your supposedly better solution as an alternative and let it win on its merits".
Yup. I could improve on systemd's logging by just piping to /dev/null. Much faster.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
And YOU sir should wake the fuck up and quit being a pedantic douchenozzle! Users ARE being banned, threads erased, discussion blocked...does this sound like "a free and open OS" to you? What we are seeing is the typical behavior of APPLE not the behavior of Linux, in fact MSFT isn't as heavy handed in their forums as RH and the new Red Debian are!
So YOU go sit in a corner and jerk off to dangling participles but mark my words the Linux users WILL have no say, systemd WILL be forced down your throat, it WILL continue to take over critical functions to become the SVCHost of Linux, and all of this will be for the benefit to Red Hat ONLY. I'd say its pretty clear who now owns Linux and it sure as fuck ain't the users, its Red hat. Hell Windows users have more say now in the direction of the OS than Linux users do and that is just sad.
ACs don't waste your time replying, your posts are never seen by me.
I kind of agree with your posts in this thread that I read. But this post of your takes the cake for a post most unrelated to it's parent post, probably for the month. Congratulations.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
Yeah I going to really miss my bash script based terminal... these new binary terminals are really terrible. /sarcasm
Uh. We are talking about error logging, inter-process communication, etc.
"Flyin' in just a sweet place,
Never been known to fail..."
OK sweety, it's alright. Mumma will come with your pacifier in a minute.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
powershelld
UNIX was a 'small, lightweight' system with small utilities surrounding it to perform specific tasks. Not a large monolithic monstrosity? It may be time to look into BSD instead of Linux as a stable ecosystem. Change for a reason that is well demonstrated is one thing, change for the sake of change is another. This sounds like change looking for a reason to be.
... "When you pry the source from my cold dead hands."
(we've recently moved from a RH contract to an Oracle Linux contract, and Oracle will bend over backwards to keep us happy..)
- This sig deliberately left blank. Nothing to see, move along.
You're willfully ignoring the objection that was made here. Now that they've decided to use systemd, in the future Gnome can't be used without it. So regardless of who decided to use systemd, once you use systemd there's no way back and any upstream project is forever bound to systemd.
The mountains of madness have many little plateaus of sanity - Terry Pratchett.
Are you sayning pipes, domain sockets or shared memory is dependent on the in-kernel VTY code ?
Note WILL, in capital letters. Typically used to denote future tense.
So you're saying it isn't now, in which case why would it be in the future?
Emphasis mine, on "has" and "existed". Typically used for past tense.
So what changed now such that created a dependence on Red Hat that didn't exist before?
You don't evaluate your server requirements or pay the maintainance for us, so you know nothing about what is suitable for us.
And the same for many other use cases, which you ignorantly choose to take notice of.
Well, if you don't want to use systemd, then don't. As you point out, it might not be needed in your case.
However, most likely whoever maintains your distro isn't going to want to maintain a configuration just for people who don't use containers. It makes more sense from their standpoint to come up with the most general solution possible, and using systemd doesn't really exclude anybody on Linux. So, you may end up having to use more of a niche distro or roll your own, which probably isn't your goal.
And your sysadmin practices are based on the assumption that VMs are the same as containers.
When did I ever make that claim?
I stated that the boot time was very useful on containers. Somebody else pointed out that he never reboots his VMs. I then pointed out that it is very helpful to be able to quickly reboot containers.
So, what is your point? That you'd rather not run systemd on your servers? Then don't. But, don't expect somebody to spend a lot of time maintaining a distro that doesn't work well on containers, as opposed to one that works well on containers and bare metal.
When Linux was first introduced, multiple VTs were revolutionary compared to MS-DOS. They have hardly changed since then. Now with move to user space, it would be much easier for anyone, including myself, to innovate. Multiple selection? Support for graphics embedded in command line stream? This has just become much more practical to implement.
I assume you're mainly a Windows user who use Linux only for the non-graphical (server, etc) stuff. Systemd is bad but it's not the Metro of Linux. The Metro of Linux would be either Unity3D (Canonical/Ubuntu) or Gnome Shell (Fedora/Redhat). These are both GUIs analogous to Metro. The rise in the popularity of Ubuntu derivative LinuxMint can be attributed to its use of its own more traditional looking desktop environments (either Cinnamon or Mate) in place of Unity. So there's clearly been a an anti-Metro-like pushback in that area.
Systemd is something else. Most desktop users probably won't notice it. And that's what makes it worse. A Systemd bug is going to be way nastier than Shellshock.
It is YOU sir that is changing the thread by acting like a third grade douchnozzle playing a game of "may I versus can I" and it seems that nobody else had ANY trouble reading the post, ONLY YOU give a rat's ass about pedantic horseshit so please, do the world a favor and get lost.
ACs don't waste your time replying, your posts are never seen by me.
No I'm saying your reading comprehension is shitty.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
This is the second time you've interjected into a discussion between myself and somebody else crying that you can't understand, yet you are the only person who cannot understand. The language is not complex which is why nobody else - including the people actually having the discussion - is having any trouble with it except you, repeatedly, so obviously it's your problem.
Doesn't have a damned thing to do with Windows or binary files, it has to do with the fact that Debian has been made Red hat's bitch by way of ex RH and Ubuntu employees taking over the board. For those that want to know what systemd is REALLY about its about cloud computing, specifically RH is pushing cloud computing like mad and systemd is gonna end up being a "SVCHost" for Linux dedicated to managing cloud computers.
This is one time me and the FOSSies are actually on the same page, as just like windows 8 was forced from on high and gave the users a big fat greasy finger so too is systemd being pushed by corporate with exactly zero fucks given about what the end users want. Ironically despite all this "empower the user" talk Linux has always had this is one case where Windows users had more power thanks to the ability to vote with their dollars, thus getting Win 8 shitcanned in favor of a much saner and nicer Win 10. But this does not mean that all hope is lost in Linux land, it just means you are gonna have to organize and SCREAM BLOODY MURDER and refuse to take this bullshit. You especially have to organize all the volunteer coders and get them to walk away, because losing all that free labor and forcing Red Hat and friends to pay for every single dime's worth of work is the ONLY way most of you can hit 'em in the pocketbook. those of you that run non cloud based servers can of course tell them you will no longer use their products but considering how much time and money you have invested in your servers I really don't see that happening.
Finally you need a rally cry, something simple and catchy and on message to focus the narrative and rally the troops, a "fuck beta" for systemd if you will. And since old Hairy will ALWAYS stand for the users allow me to give you one as a show of solidarity in your plight. Its simple, concise, on message, and sums up in a single simple sentence WTF is wrong with systemd..
SYSTEMD...Its the Metro of Linux!
As I see it for the corporate world, the server in the cloud is the way to go. It is just like outsourcing the data centre to IBM, CGI or other operations organization. It will be cheaper, it will not require a diesel generator and a computer wing, or expensive system admins or rooms of backup tapes and those couriers picking up and returning backups daily.
The server is going to be an appliance. Only if you work for the cloud company will you retain your career in Linux
Leslie Satenstein Montreal Quebec Canada
It's tons of new code, tons of attack surface, tons of bullshit.
I take it then you are not admin - or a lowly paid one.
"new code" + "tons of bullshit" = reasons to ask management for more training, more hardware, larger IT budget. Then learn some the new obscure POS and become literally irreplaceable. Every new release - another new obscure POS - and people get a chance at a nice career.
And most importantly, since RH rewrites all the stuff every time, you do not actually need the Linux experience per se. Sometimes it is even counterproductive, since RH likes to patch stuff in some freaky way. It was for quite some time now with solutions where RH products were used: "People with extensive Linux experience don't need to apply."
Just like it was in the good ol' UNIX times.
All hope abandon ye who enter here.
However, most likely whoever maintains your distro isn't going to want to maintain a configuration just for people who don't use containers.
Which distro? Ubuntu, Debian, Fedora, Arch or Gentoo? Is there a reference for any of them dropping non-container support?
So by "most likely" you mean "barely"? Or by "whoever" you mean "a minority"?
I wasn't saying that distros are going to drop non-container support. I said that they aren't going to support a different configuration just for non-container users. All of these distros support systemd, which works fine whether you use containers or not.
But, don't expect somebody to spend a lot of time maintaining a distro that doesn't work well on containers, as opposed to one that works well on containers and bare metal.
Also, don't expect everyone to spend a lot of time maintaining a distro that's fooled by propagandists who boasts about a product that brings a handful of badly implemented improvements while breaking half of everything that already work well, and refuse to fix problems in the product which lead to the breakage :)
I look forward to the obvious impending demise of Debian, Ubuntu, RHEL, CentOS, Fedora, Gentoo, Arch, Mint, and just about every other Linux distro around other than Slackware (which as far as I'm aware is the only big one that hasn't implemented systemd). Well, Google hasn't announced any plans to move away from Upstart on Chromebooks but I'm sure that is just a matter of time - other than having the best-selling laptops around I'm not sure how those matter. Maybe they'll switch from Upstart to a hodge-podge of traditional bash scripts.
You should start your own Linux distribution company founded on a systemd-free platform. It sounds like you'll have thousands of seasoned linux contributors lining up at your doors to help build your product, and you can fork Debian or whatever to start with.
It is you who concludes "has ever existed" from "will". Not me.
Bingo Dictionary - Pragmatist, n. A myopic idealist.
Ok, when are you starting to read a post before replying to it?
Bingo Dictionary - Pragmatist, n. A myopic idealist.
In 20 years of dealing with plain text unix log files, I am yet to have corruption in them, certainly not to the point where I could not view the logs at least partially.
The fact that these logs are getting corrupted most likely IS due to systemd, the developers simply don't give a fuck. "Assuming the corruption came from another source" is exactly the problem.
Also, what do you do if files are corrupted? You attempt to at least retrieve partial contents. Log files contain valuable information. Or we would not bother logging it!
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
Right as it should, but that's ok if you don't understand why.
>Sys-V is an utter shambles when automating that many machines.
That's funny it works for my broad spectrum workloads and systemd does not. Why do I not have a choice?
Funny I thought you did, since if you have "broad" workloads you probably have heavily customized a bunch of sys-v init scripts to boot up, and your upgrades are carefully planned affairs with tests, regression testing and a lot of preparation. In which case, why do you care? You're not blindly upgrading distros in the first place, so you're free to do what you want.
That, and there are the corruptable binary logs, the solution to which in the bug report is to "just delete them" and the bug has been closed as won't fix. Sorry, but if this is the resposne to journal corruption rather than finding out WHY the journals get corrupted and fixing the fuckign problem, then i do not want that in control of my logfiles.
Log files are unimportant to the people who are writing and advocating for SystemD.
Logs files are criminally important to those who have to report to regulatory agencies and otherwise important to those who manage systems for others (businesses and governments). None of those agencies will accept missing, incomplete, or corrupted log files during an investigation.
The people writing this software need to get serious or get out. Logs are of vital importance.
It seems like I am surrounded by inexperienced children who have no adults to discipline them.
People who seem to think log files are so important sure seem to spend a lot of time not learning about log files in systemd before writing angry internet posts about them. Otherwise they'd know that you can forward text logs to any system you want with journald, and avoid the binary log format entirely. Which is information one can find out from say, man pages, Google or any of the sources I presume they consult when configuring such important log file storage systems.
I mean, I also assume they're up to speed on problems with syslog, like how by default the whole protocol is UDP and will drop packets freely?
Most end users don't care about the init system one way or another, since most end users don't ever mess with it. On the other hand, every end user was forced to mess with Metro. That's the difference.
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
The Linux kernel is monolithic in design but modular in architecture.
Systemd is the opposite and is the wrong way to approach it.
Also in regards to kernel development, monolith is a overloaded term.
A microkernel shoves as much crap into user space as it can, a monolithic kernel does not. It doesn't mean its architecture has to be monolithic, the Linux kernel is proof of that.
For all of the different processes in systemd, they might as well be shoved in one process because that is exactly the effect that it gives having all its processes tightly coupled to each other.
If people could pull out BS like the logger and login and seamlessly add their own sane processes, the complaints would mostly disappear. Instead, you have to take the binary logger and use that buggy mess to forward it to a sane text logger.
If you systemd apologists can't see how any of this is poor design and detrimental to linux you simply lack the technical knowledge and should stop white knighting for that idiot poettering.
You actually think that adding bloat is a good thing? Using journald to pipe logs to rational and sane logging systems is a solution in your world?