systemd: Yet Another Init Replacement

lymphatik · 2011-04-24 14:05:23

I also have a question what is the point of @.service for systemd ? Actually found the explanation in the man page.

Last edited by lymphatik (2011-04-24 14:42:13)

Stebalien · 2011-04-24 17:23:14

@lymphatik
I use the following unit for laptop-mode and it seems to work:

[Unit]
Description=Service to extend battery life.
After=acpid.service
Wants=acpid.service

[Service]
RemainAfterExit=yes
ExecStartPre=/bin/install -d /var/run/laptop-mode-tools ; /bin/touch /var/run/laptop-mode-tools/enabled
ExecStart=/usr/sbin/laptop_mode auto

ExecStop=/usr/sbin/laptop_mode stop
ExecStopPost=/bin/rm -f /var/run/laptop-mode-tools/enabled

ExecReload=/usr/sbin/laptop_mode auto force

[Install]
WantedBy=multi-user.target

lymphatik · 2011-04-24 19:02:21

@Stebalien thanks seems to work flawlessly.

pogeymanz · 2011-04-25 00:52:41

EDIT: I just needed to do some more reading. Move along. Nothing to see here.

Last edited by pogeymanz (2011-04-25 19:21:42)

ChoK · 2011-04-25 20:10:51

pankajmore wrote:

on boot i get this

<30>systemd-fsck[772]: /dev/sda1: clean, 400835/2501856 files, 3939069/10000454 blocks
<30>systemd-fsck[2424]: /dev/sda4: Superblock last write time is in the future.
<30>systemd-fsck[2424]: (by less than a day, probably due to the hardware clock being incorrectly set). FIXED.
<30>systemd-fsck[2424]: /dev/sda4: clean, 717737/26402816 files, 68396908/105595245 blocks
<30>systemd-fsck[2428]: Reiserfs super block in block 16 on 0x802 of format 3.6 with standard journal
<30>systemd-fsck[2428]: Blocks (total/free): 5000224/1279347 by 4096 bytes
<30>systemd-fsck[2428]: Filesystem is clean

time changes at each boot , any idea how to fix it .... I am using UTC and no windows

Had the same problem. Try disabling the hwclock-load.service and report back. Are you using ntpd ?

Last edited by ChoK (2011-04-25 20:11:17)

mrmylanman · 2011-04-25 22:40:34

Holy cow, I just got systemd set up (barely, I didn't do hardly any configuration yet) and my boot time went down from 14.5 seconds to 10.3 seconds. I have an SSD so it's nice to see that paying off now... haha. Wish I could help you out ajk I have a laptop so I'm using network manager and it still works fine.

darkraven · 2011-04-27 01:35:18

systemd works fine with the arch official kernel.
But when work with my customized kernel, i have seen some error when trying to shutdown the computer.
It takes quite long time to poweroff the computer, and kernel message says something like 'ata1.00: exception Emask....' 'ata1.00: cmd....'.
This means something trys to read the disk while the disk is actually stopped.
And I have no such problem when using sysvinit with the customized kernel or using systemd with arch official kernel.
I've selected all required kernel options listed in archwiki.

I've tried the latest zen-kernel and vanilla kernel 2.6.39-rc4

EDIT: I build some debug info into libata, the failed cmd is FPDMA READ QUEUED. But I don't know which process did it

Last edited by darkraven (2011-04-27 05:18:48)

mrmylanman · 2011-04-27 11:03:34

I just noticed something similar as well. I'm using the 2.6.38-pf kernel, and after I tell it to shutdown on the terminal I see network manager stopping, and then acpid stopping (sometimes network manager stops after acpid), and then it just hangs, although I didn't let it sit TOO long (maybe a couple minutes).

I'll try and find some more specific information when I can. Before this I just put the laptop to sleep rather than shutting it down so I never noticed.

darkraven · 2011-04-27 14:10:13

mrmylanman wrote:

I just noticed something similar as well. I'm using the 2.6.38-pf kernel, and after I tell it to shutdown on the terminal I see network manager stopping, and then acpid stopping (sometimes network manager stops after acpid), and then it just hangs, although I didn't let it sit TOO long (maybe a couple minutes).
I'll try and find some more specific information when I can. Before this I just put the laptop to sleep rather than shutting it down so I never noticed.

I don't think we are experiencing the same bug....

mrmylanman · 2011-04-27 17:18:48

That could very well be the case. I did notice fine behavior with the stock kernel with systemd, but when I moved to kernel26-pf it hangs when I go to shutdown indefinitely (I let it sit for 15 minutes this morning). The output is different every time I go to shutdown, I noticed as well, it just usually happens to hang when acpid and network manager are stopping... I'll look into it some more, or maybe just go back to the stock kernel or try another one maybe.

darkraven · 2011-04-28 16:07:21

mrmylanman wrote:

That could very well be the case. I did notice fine behavior with the stock kernel with systemd, but when I moved to kernel26-pf it hangs when I go to shutdown indefinitely (I let it sit for 15 minutes this morning). The output is different every time I go to shutdown, I noticed as well, it just usually happens to hang when acpid and network manager are stopping... I'll look into it some more, or maybe just go back to the stock kernel or try another one maybe.

My kernel hangs when everything is stopped, and reboot() called.
the kernel compliants about something reading the stopped disk.

Now I found the problem only occur when boot without initrd, otherwise everything is fine.

mrmylanman · 2011-04-29 00:42:03

Yeah I am not sure what is going on with mine. I tried checking out the logs but I couldn't find any relevant information. I am going to try some other kernels and see what my luck is there.

Edit:

Got it; I should have read the Wiki closer.

Because of the really fast boot times (with the kernel26-lqx kernel it's about 9 and a half seconds now) I didn't notice the output:

/etc/mtab is not a symlink or not pointing to /proc/self/mounts. This is not supported anymore. Please make sure to replace this file by a symlink to avoid incorrect or misleading mount(8) output.

So, I didn't think to consult the Wiki, which suggests the following command to fix the issue:

ln -fs /proc/self/mounts /etc/mtab

Did so, and it fixed the problem. So I'm just an idiot :-)

Last edited by mrmylanman (2011-04-29 01:21:48)

eworm · 2011-04-29 12:32:39

Ok, found the next stupid problem... cgroups are not all good.
If I connect to a network within X the dhcpcd process belongs to the login manager's cgroup. So 'systemctl restart slim.service' kills my network... Can I avoid this?
The process relationships: slim -> xfce -> terminal -> su -> tmux -> netcfg

stefano.facchini · 2011-04-29 13:59:34

Is there any native configuration option equivalent to the TIMEZONE variable set in rc.conf? The systemd wiki page says only that TIMEZONE is honored.

denisfalqueto · 2011-04-29 14:25:31

eworm wrote:

Ok, found the next stupid problem... cgroups are not all good.
If I connect to a network within X the dhcpcd process belongs to the login manager's cgroup. So 'systemctl restart slim.service' kills my network... Can I avoid this?
The process relationships: slim -> xfce -> terminal -> su -> tmux -> netcfg

You could use wicd and it would be independent X. Even though the graphical interface is used in X, the backing server is independent and will keep the connection between X restarts.

eworm · 2011-04-29 14:41:41

denisfalqueto wrote:

You could use wicd and it would be independent X. Even though the graphical interface is used in X, the backing server is independent and will keep the connection between X restarts.

But I like netcfg.
Anyway... there's a lot of software that forks itself to background to run as a daemon. How to get it out of the X cgroup?

denisfalqueto · 2011-04-29 15:10:34

eworm wrote:

Anyway... there's a lot of software that forks itself to background to run as a daemon. How to get it out of the X cgroup?

I think that the wiki about netcfg explains some forms of running it on the boot process.

dakt · 2011-04-29 15:59:53

Article from LWN:
http://lwn.net/Articles/440636/

taylorchu · 2011-04-29 23:46:28

it looks great. Will arch use this by default?

unikum · 2011-04-30 06:16:44

@taylorchu: in the near future is unlikely, because it yet unstable.

frankieboy · 2011-04-30 09:12:27

Hi!

I've tried systemd from the community repo again, what I've found that both on my laptop and on my home computer, the mount of the boot partition fails with a time-out, and if I gave the root password, I can manually mount the boot partition, why is that?
In my fstab, there is nothing specific:
UUID=something /boot ext4 defaults,noatime,nodiratime 1 2
On the home machine, after the timeout, it fails to mount /home and another partition mounted under home, it says dependency error.
On my laptop, networkmanager with systemd starts randomly, booting once everything is ok, on the other time, it fails.

Any help appreciated.

Thank you.

adrianx · 2011-04-30 09:24:54

I had problems with networkmanager after updating to systemd-arch-units-20110429-1. What solved it for me was adding "After=syslog.target" back to networkmanager.service (like it was in systemd-arch-units-20110411-1)

[Unit]
Description=Network Manager
After=syslog.target

[Service]
Type=dbus
BusName=org.freedesktop.NetworkManager
ExecStart=/usr/sbin/NetworkManager --no-daemon
StandardOutput=syslog

[Install]
WantedBy=multi-user.target

Glaucous · 2011-04-30 13:53:53

Thank you, I had the very same problem.

MartinZ · 2011-04-30 18:26:57

adrianx wrote:

I had problems with networkmanager after updating to systemd-arch-units-20110429-1. What solved it for me was adding "After=syslog.target" back to networkmanager.service (like it was in systemd-arch-units-20110411-1)

I'm also having problems with latest systemd-arch-units, but unfortunately are not fixed this way. The symptoms are:

1. KDE takes ages to load and is very unresponsive (though the boot is as fast as usual)
2. Networkmanager does not work (even with the precedent fix)
3. Shutting the laptop down lets it somehow hangup at a console screen. I can change through tty's, but can't write nor login.

Downgrading systemd-arch-units to 20110411-1 solves everything.
I really don't know how to give more info, sorry for the poor report.

falconindy · 2011-04-30 19:37:38

I suggest switching to rsyslog, until syslog-ng can get its act together (and implement proper socket activation and takeover). This should fix a large number of problems that people have been experiencing recently wrt hanging at shutdown (and probably bootup as well).

community/rsyslog includes its own service file, so its simply a matter of HUP'ing systemd and enabling the new logger.

Arch Linux

#451 2011-04-24 14:05:23

Re: systemd: Yet Another Init Replacement

#452 2011-04-24 17:23:14

Re: systemd: Yet Another Init Replacement

#453 2011-04-24 19:02:21

Re: systemd: Yet Another Init Replacement

#454 2011-04-25 00:52:41

Re: systemd: Yet Another Init Replacement

#455 2011-04-25 20:10:51

Re: systemd: Yet Another Init Replacement

#456 2011-04-25 22:40:34

Re: systemd: Yet Another Init Replacement

#457 2011-04-27 01:35:18

Re: systemd: Yet Another Init Replacement

#458 2011-04-27 11:03:34

Re: systemd: Yet Another Init Replacement

#459 2011-04-27 14:10:13

Re: systemd: Yet Another Init Replacement

#460 2011-04-27 17:18:48

Re: systemd: Yet Another Init Replacement

#461 2011-04-28 16:07:21

Re: systemd: Yet Another Init Replacement

#462 2011-04-29 00:42:03

Re: systemd: Yet Another Init Replacement

#463 2011-04-29 12:32:39

Re: systemd: Yet Another Init Replacement

#464 2011-04-29 13:59:34

Re: systemd: Yet Another Init Replacement

#465 2011-04-29 14:25:31

Re: systemd: Yet Another Init Replacement

#466 2011-04-29 14:41:41

Re: systemd: Yet Another Init Replacement

#467 2011-04-29 15:10:34

Re: systemd: Yet Another Init Replacement

#468 2011-04-29 15:59:53

Re: systemd: Yet Another Init Replacement

#469 2011-04-29 23:46:28

Re: systemd: Yet Another Init Replacement

#470 2011-04-30 06:16:44

Re: systemd: Yet Another Init Replacement

#471 2011-04-30 09:12:27

Re: systemd: Yet Another Init Replacement

#472 2011-04-30 09:24:54

Re: systemd: Yet Another Init Replacement

#473 2011-04-30 13:53:53

Re: systemd: Yet Another Init Replacement

#474 2011-04-30 18:26:57

Re: systemd: Yet Another Init Replacement

#475 2011-04-30 19:37:38

Re: systemd: Yet Another Init Replacement

Board footer