systemd: Yet Another Init Replacement

WorMzy · 2011-05-26 01:49:46

reflexing wrote:

falconindy wrote:

Sounds like /etc/mtab isn't a symlink to /proc/self/mounts. If that isn't the case, you'll need to post your /etc/fstab.

Of course it is a symlink. Here is my /etc/fstab:

# 
# /etc/fstab: static file system information
#
# <file system>        <dir>         <type>    <options>          <dump> <pass>
devpts                 /dev/pts      devpts    defaults            0      0
shm                    /dev/shm      tmpfs     nodev,nosuid        0      0
/dev/mapper/arch-home /home ext4 defaults,user_xattr 0 1
/dev/mapper/arch-media /mnt/media xfs defaults 0 1
/dev/mapper/arch-root / ext4 defaults 0 1
/dev/mapper/arch-swap swap swap defaults 0 0
/dev/mapper/arch-var /var reiserfs defaults 0 1
/dev/mapper/arch-boot /boot ext2 defaults 0 1
none /proc/bus/usb usbfs devgid=108,devmode=664 0 0

On a (possibly) unrelated note, you've used fs_passno=1 (the last column) for five of your partitions. You should only use "1" for the root partition; use "2" for everything else. (read fstab's manpage for more details)

franz1789 · 2011-05-26 09:02:54

and what about my problem? I'm sorry to ask, but as now I can still access my arch..

WorMzy · 2011-05-26 10:56:32

You'll have to wait for falconindy to get back to you on that. In the mean time, you may want to check your fstab to see if you've made the same mistake as reflexing.

franz1789 · 2011-05-26 11:33:24

here's my fstab:

# 
# /etc/fstab: static file system information
#
# <file system>        <dir>         <type>    <options>          <dump> <pass>
devpts                 /dev/pts      devpts    defaults            0      0
shm                    /dev/shm      tmpfs     nodev,nosuid        0      0

#/dev/cdrom             /media/cd   auto    ro,user,noauto,unhide   0      0
#/dev/dvd               /media/dvd  auto    ro,user,noauto,unhide   0      0
#/dev/fd0               /media/fl   auto    user,noauto             0      0

/dev/sda1 /windows ntfs-3g defaults,locale=it_IT.utf8 0 0
/dev/sda2 / ext4 defaults,noatime 0 1
/dev/sda3 swap swap defaults 0 0
/dev/sda5 /home ext3 defaults,noatime 0 2

think it's ok.. anyway, I'll wait for help.. thank you

falconindy · 2011-05-26 11:58:27

I can't help if you can't help yourself. I've mentioned adding systemd.log_level=debug to your bootloader to get more detail as to why you're crashing into the rescue shell.

franz1789 · 2011-05-26 12:18:43

falconindy wrote:

I can't help if you can't help yourself. I've mentioned adding systemd.log_level=debug to your bootloader to get more detail as to why you're crashing into the rescue shell.

and as I asked before: should I add this line to the kernel line in grub?

falconindy · 2011-05-26 12:29:11

As mentioned in the manpage for systemd: yes, this is a flag that systemd reads from the kernel command line in your bootloader.

falconindy · 2011-05-26 18:49:00

systemd 28 is out. hwclock-{load,save}.service is now extinct:

http://lists.freedesktop.org/archives/s … 02526.html

franz1789 · 2011-05-27 11:35:08

Here's what I get when adding that line:

Failed to issue method call: Transaction is destructive

and then the same story: I give root password, fsck says all ok, systemctl default return the above message..

falconindy · 2011-05-27 12:00:16

There's an awful lot more output than just that when you're using loglevel debug. Based on that singular message alone, I can't even begin to speculate as to what might be causing it.

franz1789 · 2011-05-27 12:28:42

I'll look if I wrote it right (no doubt since I wrote it and checked it ten times).. sad to see I'm the only one having this problem..

falconindy · 2011-05-27 13:22:45

One of the first rules in debugging: the error itself isn't necessarily the most important part, but what leads up to it.

franz1789 · 2011-05-27 15:53:28

I'd be proud of helping debug, but I don't really know what to do. In addition to dmesg and logs in /var/log, is there a specific systemd log file? the fact is that I'm new in systemd (and systemd is quite new itself) so for me it's difficult to find files and solve problems, and while being well documented, I'm perhaps the first one having this problem..

P.S. I've found this, noone answered..

Wey · 2011-05-27 16:00:57

franz, make sure to remove any "quiet"-option from grubs kernel line as well.

franz1789 · 2011-05-27 16:06:17

Wey wrote:

franz, make sure to remove any "quiet"-option from grubs kernel line as well.

This is quite possible, and the fact that I remember that word in my kernel line make me feel like an idiot.. as soon as I'll be home, I'll post some more information about my problem

franz1789 · 2011-05-28 18:49:41

solved it. Just used systemctl --failed and saw that windows.mount, for some reason was always failing.. then I simply commented the line of windows partition in fstab and it start working again.. remains a mistery why this happened, since it just says that partitions seems to be already mounted/used/open..

hoschi · 2011-05-29 17:05:29

Hello!
When I use the terminal or Gnome3 I can't reboot reliable, the computer just "hangs" and displays the last output. When I use "systemctl reboot" it seems to work?
I already read this - https://wiki.archlinux.org/index.php/Sy … _reboot.3F

But I am using rsyslog. I found this as mentioned one page ago, but I didn't removed anything (except syslog-ng).
Boot, halt, pm-suspend and everything else works perfect, but "reboot" is not reliable.
Any idea?

Found this:
-> http://lists.suse.com/opensuse-bugs/201 … 02041.html

I have a self-compiled kernel, with "transparent Huge Page" support (without using hugetblfs).
How can I tell systemd I don't want this?!

Last edited by hoschi (2011-05-29 17:22:33)

blamed · 2011-05-29 19:55:47

Hi.
How can I find which services are enabled?
Not by looking in /etc/systemd/system/[some_dirs]. It's.. Uncomfortable, becouse there are many folders.

With sysV i can simply look at DAEMONS array. Is there something like that? Something like "systemctl list-enabled"?

ethail · 2011-05-29 20:09:05

@blamed:

https://wiki.archlinux.org/index.php/Systemd

Listing running services
$ systemctl

litemotiv · 2011-05-29 20:15:03

blamed wrote:

Hi.
How can I find which services are enabled?
Not by looking in /etc/systemd/system/[some_dirs]. It's.. Uncomfortable, becouse there are many folders.
With sysV i can simply look at DAEMONS array. Is there something like that? Something like "systemctl list-enabled"?

Try:

$ systemctl

oops too late..

Last edited by litemotiv (2011-05-29 20:15:29)

James_Eder · 2011-06-01 15:48:30

If it's not too much trouble can we have /lib/systemd/arch-persistent-settings changed from saying "# Autogenerated by systemd, do not edit" to something like "# Autogenerated by arch-persistent-settings. Do not edit."?

I had been looking through systemd's documentation and finding nothing because of it. Such a change would make it more obvious and less confusing.

BasT · 2011-06-01 19:25:42

So I read systemd for Administrators, Part VII - The Blame Game and decided to try and improve boot times even further on my two systems.

On my netbook udev-trigger.service seem to take an extraodinary amount of time (on the desktop pc it's just 319ms).

$ systemd-analyze blame
  2121ms udev-trigger.service
   261ms systemd-vconsole-setup.service
   200ms rsyslog.service
   184ms networkmanager.service
[...]

On the netbook I use a custom compiled kernel with pretty much all required drivers compiled into the kernel. Would it be safe to remove the udev-trigger.service from the boot process?
Or analyse somehow why it's taking so long?

On the dektop pc mounting my two ntfs-harddrives takes rather long:

$ systemd-analyze blame
  1089ms mnt-data3.mount
   722ms mnt-data2.mount

I tried adding comment=systemd.automount to their mount options but that actually increased their time by 200ms each! What's up with that? Shouldn't they then only be mounted on demand rather than holding up the boot process?

Edit:
Also, on the netbook $ systemctl tells me:

$ systemctl
var-lock.mount            loaded failed failed       Lock Directory

Is that a problem? How might I fix it?

Last edited by BasT (2011-06-01 19:29:39)

falconindy · 2011-06-01 19:54:41

BasT wrote:

So I read systemd for Administrators, Part VII - The Blame Game and decided to try and improve boot times even further on my two systems.
On my netbook udev-trigger.service seem to take an extraodinary amount of time (on the desktop pc it's just 319ms).
$ systemd-analyze blame
  2121ms udev-trigger.service
   261ms systemd-vconsole-setup.service
   200ms rsyslog.service
   184ms networkmanager.service
[...]
On the netbook I use a custom compiled kernel with pretty much all required drivers compiled into the kernel. Would it be safe to remove the udev-trigger.service from the boot process?
Or analyse somehow why it's taking so long?

udev-trigger is what loads modules for your hardware. You're free to remove whatever you want as long as you understand the implications. If you want to see what it's actually doing, feel free to boot to rescue.target and run 'udevadm monitor' alongside the 2 udevadm trigger calls. I recall some discussion about speeding up trigger on the udev mailing list, but I can't recall specifics offhand.

On the dektop pc mounting my two ntfs-harddrives takes rather long:
$ systemd-analyze blame
  1089ms mnt-data3.mount
   722ms mnt-data2.mount
I tried adding comment=systemd.automount to their mount options but that actually increased their time by 200ms each! What's up with that? Shouldn't they then only be mounted on demand rather than holding up the boot process?

You need to add 'noauto' to your mountpoints with comment=systemd.automount. Some FSs take longer to sync up with the kernel than others. ext4 is (much) faster than say, btrfs, in this regard.

Edit:
Also, on the netbook $ systemctl tells me:
$ systemctl
var-lock.mount            loaded failed failed       Lock Directory
Is that a problem? How might I fix it?

Offhand, no idea. Is /var a separate partition?

Keep in mind that the blame list doesn't properly represent the parallel nature of your bootstrap. use the 'plot' function to get a better idea of where the bottlenecks are.

Last edited by falconindy (2011-06-01 19:56:26)

sonay · 2011-06-03 15:55:04

While booting I get a

"hda-codec: no NID mapping control independent hp:0:0"

message, but everything works fine. Google search did not return any useful results, What do you think? I am using alsa.service.

lspci -vvv for my audio device is

00:1b.0 Audio device: Intel Corporation N10/ICH 7 Family High Definition Audio Controller (rev 02)
Subsystem: CLEVO/KAPOK Computer Device 1100
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0, Cache Line Size: 32 bytes
Interrupt: pin A routed to IRQ 44
Region 0: Memory at f0400000 (64-bit, non-prefetchable) [size=16K]
Capabilities: [50] Power Management version 2
Flags: PMEClk- DSI- D1- D2- AuxCurrent=55mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
Capabilities: [60] MSI: Enable+ Count=1/1 Maskable- 64bit+
Address: 00000000fee0300c Data: 4181
Capabilities: [70] Express (v1) Root Complex Integrated Endpoint, MSI 00
DevCap: MaxPayload 128 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
ExtTag- RBE- FLReset-
DevCtl: Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
MaxPayload 128 bytes, MaxReadReq 128 bytes
DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr+ TransPend-
LnkCap: Port #0, Speed unknown, Width x0, ASPM unknown, Latency L0 <64ns, L1 <1us
ClockPM- Surprise- LLActRep- BwNot-
LnkCtl: ASPM Disabled; Disabled- Retrain- CommClk-
ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
LnkSta: Speed unknown, Width x0, TrErr- Train- SlotClk- DLActive- BWMgmt- ABWMgmt-
Capabilities: [100 v1] Virtual Channel
Caps: LPEVC=0 RefClk=100ns PATEntryBits=1
Arb: Fixed- WRR32- WRR64- WRR128-
Ctrl: ArbSelect=Fixed
Status: InProgress-
VC0: Caps: PATOffset=00 MaxTimeSlots=1 RejSnoopTrans-
Arb: Fixed- WRR32- WRR64- WRR128- TWRR128- WRR256-
Ctrl: Enable+ ID=0 ArbSelect=Fixed TC/VC=01
Status: NegoPending- InProgress-
VC1: Caps: PATOffset=00 MaxTimeSlots=1 RejSnoopTrans-
Arb: Fixed- WRR32- WRR64- WRR128- TWRR128- WRR256-
Ctrl: Enable+ ID=1 ArbSelect=Fixed TC/VC=80
Status: NegoPending- InProgress-
Capabilities: [130 v1] Root Complex Link
Desc: PortNumber=0f ComponentID=02 EltType=Config
Link0: Desc: TargetPort=00 TargetComponent=02 AssocRCRB- LinkType=MemMapped LinkValid+
Addr: 00000000fed1c000
Kernel driver in use: HDA Intel
Kernel modules: snd-hda-intel

Btw I am trying to move to a native system, but I don't know which modules I need to load. For now, I still depend on rc.conf's auto load modules. How can I proceed?

Finally, does this system have to fsck every time I boot? I use comment=systemd.automount for my data disk. But what should I do for /?

Last edited by sonay (2011-06-03 16:03:39)

sonay · 2011-06-03 16:43:40

sonay wrote:

Btw I am trying to move to a native system, but I don't know which modules I need to load. For now, I still depend on rc.conf's auto load modules. How can I proceed?

I used

lspci -vvv | grep "Kernel modules"

created the corresponding *.conf files and moved rc.conf just to check. And everything seems to work, was that a healthy method?

Arch Linux

#676 2011-05-26 01:49:46

Re: systemd: Yet Another Init Replacement

#677 2011-05-26 09:02:54

Re: systemd: Yet Another Init Replacement

#678 2011-05-26 10:56:32

Re: systemd: Yet Another Init Replacement

#679 2011-05-26 11:33:24

Re: systemd: Yet Another Init Replacement

#680 2011-05-26 11:58:27

Re: systemd: Yet Another Init Replacement

#681 2011-05-26 12:18:43

Re: systemd: Yet Another Init Replacement

#682 2011-05-26 12:29:11

Re: systemd: Yet Another Init Replacement

#683 2011-05-26 18:49:00

Re: systemd: Yet Another Init Replacement

#684 2011-05-27 11:35:08

Re: systemd: Yet Another Init Replacement

#685 2011-05-27 12:00:16

Re: systemd: Yet Another Init Replacement

#686 2011-05-27 12:28:42

Re: systemd: Yet Another Init Replacement

#687 2011-05-27 13:22:45

Re: systemd: Yet Another Init Replacement

#688 2011-05-27 15:53:28

Re: systemd: Yet Another Init Replacement

#689 2011-05-27 16:00:57

Re: systemd: Yet Another Init Replacement

#690 2011-05-27 16:06:17

Re: systemd: Yet Another Init Replacement

#691 2011-05-28 18:49:41

Re: systemd: Yet Another Init Replacement

#692 2011-05-29 17:05:29

Re: systemd: Yet Another Init Replacement

#693 2011-05-29 19:55:47

Re: systemd: Yet Another Init Replacement

#694 2011-05-29 20:09:05

Re: systemd: Yet Another Init Replacement

#695 2011-05-29 20:15:03

Re: systemd: Yet Another Init Replacement

#696 2011-06-01 15:48:30

Re: systemd: Yet Another Init Replacement

#697 2011-06-01 19:25:42

Re: systemd: Yet Another Init Replacement

#698 2011-06-01 19:54:41

Re: systemd: Yet Another Init Replacement

#699 2011-06-03 15:55:04

Re: systemd: Yet Another Init Replacement

#700 2011-06-03 16:43:40

Re: systemd: Yet Another Init Replacement

Board footer