SystemD Has New Shutdown Logic, Gives Everyone CGroups

[dave-practicalclouds]

Syetemd can be quite tricky to get right!

Here's a gotcha I ran into trying to set a script to run at halt, init 0 etc...

http://www.practicalclouds.com/content/blog/1/dave-mccormick/2012-02-27/why-dont-my-systemd-shutdown-scripts-run

regards



Dave

[liam]

Phoronix: SystemD Has New Shutdown Logic, Gives Everyone CGroups

Fedora 14 was set to be the first major distribution shipping SystemD to replace SysVinit, but that ended up getting pushed back to the Fedora 15 release that will now come in May of 2011. Fortunately, for the developers behind Fedora and SystemD, this means the init replacement daemon will be in much better shape for its premiere. Lennart Poettering, the original developer of SystemD, has written about some of the recent improvements...

http://www.phoronix.com/vr.php?view=ODc5OQ

There's been discussion on the gnome-shell mailing list about lightdm as a replacement for gdm. I wonder why we even have gdm any longer?
Does anyone know if it's possible to move the desktop init to systemd (calling scripts which would initialize various components and then exiting rather than having gdm running the entire damn time)? From following Lennart's work it seems as though it would be a great alternative and without the additional overhead of a separate dm.

[Micket]

LLVM, Galium3D, responsiveness, systemd, E17 in beta 2... 2011 is going to rock! ^^,

I'd put in Btrfs in that list.

Now, if only intel showed some more interest in Gallium3D i could be more excited.

[DeepDayze]

It's (as usually) about trade-offs, more backwards compatibility - more problems having a true new and modern system and there's always be people not agreeing on particular time/design decisions.

Little new stuff - people will say it doesn't bring enough to the table to justify a new init system.
Just enough new stuff - people will say it should have had more new features since it's a new init system.
A lot of new stuff - people will say it's too complicated.

Exactly!

Tradeoffs that need to be thought thru carefully and kicked around...best implementation should win

[KDesk]

Canonical is looking at it for Ubuntu

I thought it would take more time until they realize that upstart is not such a great idea after all, and that there can be something better. Do you have more information about this? Any announcement from them?

It's spelled "systemd" not "SystemD". It's not capitalized like that anywhere in the official docs, Lennart's blog posts, or even the Wikipedia article.

You are right. From http://freedesktop.org/wiki/Software/systemd

Yes, it is written systemd, not system D or System D, or even SystemD. And it isn't system d either. Why? Because it's a system daemon, and under Unix/Linux those are in lower case, and get suffixed with a lower case d. And since systemd manages the system, it's called systemd. It's that simple. The only situation were we find it OK to use an uppercase letter in the name (but don't like it either) is if you start a sentence with systemd. On high holidays you may also spell it s?st?md. But then again, Syst?me D is not an acceptable spelling and something completely different (though kinda fitting).

[V!NCENT]

LLVM, Galium3D, responsiveness, systemd, E17 in beta 2... 2011 is going to rock! ^^,

[Ragas]

I actually think systemd will be the biggest thing for linux in 2011.

[V!NCENT]

Many people still think that sysd is just a 'bombs away' startup thing that boots faster. It's much more. Systemd improves stability and robustness. If a dbus crashes it's not a problem. Nothing using dbus needs to be restarted and you won't even know it happened.

Sysd even reduces memory footprint, reduced battery consumption and is also good for embedded.

[Ragas]

Is there any hope for improving/updating the classic sysvinit? There needs to be an init system that's easy to manage and maintain. I like how Slackware keeps its init scripts simple, and not the complex mess that's in most distros today. If there needs to be a new init system, it needs to be eay to manage and not break compatibility with the older init systems

seeing how far systemd is right now I have no urge tho have old sysvinit-style services anymore.
sure there are some missing things in this but those aren't the most important things and this init-system is so incredible you wouldn't want anything else.

Right now im using it on my play system, but I'm also thinking of compiling it for gentoo.

[cl333r]

Is there any hope for improving/updating the classic sysvinit? There needs to be an init system that's easy to manage and maintain. I like how Slackware keeps its init scripts simple, and not the complex mess that's in most distros today. If there needs to be a new init system, it needs to be eay to manage and not break compatibility with the older init systems

It's (as usually) about trade-offs, more backwards compatibility - more problems having a true new and modern system and there's always be people not agreeing on particular time/design decisions.

Little new stuff - people will say it doesn't bring enough to the table to justify a new init system.
Just enough new stuff - people will say it should have had more new features since it's a new init system.
A lot of new stuff - people will say it's too complicated.