Adding simple commands to boot under systemd

now that the rc.xxxx files are gone, where is the preferred location for adding a simple command one might like to execute upon boot, e.g. "echo 'search somewhere.somedomain' >> /etc/resolv.conf"
obviously there are several ways to do this, i'm just wondering what a systemd expert would do.

djg1971 wrote:olive: please try to think a little more abstractly. it is not for you to judge whose one-line commands are good, not good, stupid, etc.
Please read my posts more carefully. I never said that there is no valid reason to run a custom command at boot, I was just pointed out, in a part of my post, that your example was not the best one; that's all and I do not understand why it seems to affect you so much.
Anyway, forget about the bad example. I have tried to answer your original question the best as I could: to run a custom command at boot, you have to write a .service or use a general rc-local script but apparently you do not like the answer. What do you want now? I do not know if I am an expert but I have read enough of systemd to know there is no real other way: systemd take in charge every program that is run at boot via .services (or .target, etc...) files. So this is the way you have to follow to run a custom command at boot. Writing a .service file is not that difficult, maybe easier than to write a custom rc.somewhat file. Just take one existing .service file, understand what there is in it and modify it.
The post of ewaller give a method to create tempory files, often used by .service files (or the programs they launches); I was not under the impression that is what you really want.
As for the benefits of systemd, there are a few advantages. Anyway it is not me that I have decided to switch to systemd and discussing lengthy of the advantages of disadvantages of it would lead to nowhere.
Last edited by olive (2013-02-25 22:09:21)

Similar Messages

Run 'echo' command at boot.

Hey!
I am trying to run the following command at boot:
"echo 10 > /sys/class/backlight/acpi_video0/brightness"
So far I've tried in /etc/rc.local. That didn't work
Any thoughts?

I can think of a primitive way to do it, but there's probably a better (and simpler) way than this...
Edit ~/.xinitrc and add your command above the "exec ..." line:
$ nano ~/.xinitrc
echo 10 > /sys/class/backlight/acpi_video0/brightness &
exec ck-launch-session gnome-session
The "&" at the end should be probably added in, for good measure. Backgrounding this task is (probably) a good idea, else your GNOME session (or whatever DE/WM you use) may not start.
Then edit your sudoers file, so that it doesn't ask for a password:
$ sudo EDITOR=nano visudo
Defaults editor=/usr/bin/nano
root ALL=(ALL) ALL
dspider ALL=(ALL) ALL
dspider ALL=NOPASSWD: /bin/mount, /bin/umount
dspider arch=NOPASSWD: /sbin/reboot, /sbin/poweroff, /usr/bin/pacman -Syu, /bin/echo 10 > /sys/class/backlight/acpi_video0/brightness
The easier way is to use the "-b" flag from gtk-redshift (inspired by f.lux):
# pacman -S redshift
$ gtk-redshift -l 44.42:26.10 -b 0.5 -g 0.8 -t 6500:3800 -m vidmode &
Last edited by DSpider (2012-09-26 07:16:29)

SCCM 2012 - Problems running vbscript or simple command line

Hey everybody, I'm having a heck of a time trying to get SCCM 2012 to deploy some very simple Vbscripts or even a simple command line. I need to deploy some registry fixes across the company as well as a Vbscript that uninstalls some old applications.
If I run these scripts/commands outside of SCCM everything works fine, inside of SCCM it always shows "succeeded" but it doesn't actually run correct on the client.
What's confusing to me is the use of distribution points when you're running a command line that doesn't actually have any source files. For instance this is what I'm trying to run "regedit /s \\servername\folder\registryfix.reg". I've also tried
a similar bat file as well with the same result. I've tried it with a source folder specified, I've tried it using Run Program From Server or Download locally. I've gone through everything I can think of but can't get it to work.
Similarly with a vbscript I'm trying "cscript.exe /s \\servername\folder\uninstallapps.wsf" and various different settings in the package deployment.
I'm really confused here guys so any help would be greatly appreciated. Thanks

What I have found is that if you try and run the REG.EXE command line directly, either as a program or via a CMD/BAT file, the 64-bit redirection will occur. So, if you had something like this in a program or via a CMD/BAT file that ran against a 64-bit
system:
REG ADD HKLM\Software\Microsoft\Test /v Test /d Test
You would end up with this registry entry existing under:
HKLM\Software\Wow6432Node\Microsoft\Test
That's because SCCM uses the 32-bit launcher to take care of this and redirection occurs - You can see this via the "CCM32BitLauncher.log" on the client.
To work around this problem, I have done the following:
On a reference machine, create/export the registry entry I want into a *.reg file - An example export of "Test.reg" for the above mentioned registry entry would look something like this:
Windows Registry Editor Version 5.00
[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Test]
"Test"="Test"
2. Stick our "Test.reg" file in a source share, create a package with the share as the source content, and distribute it
3. Create a program for the package with a command line like this:
REG IMPORT Test.reg
This is obviously not as clean as having a command line with a package to distribute, but I find it works fine for me and gets around the redirection for registry entries.

Invalid command in boot strap mode

Hi All,
I am trying to install peoplesoft HRMS 9.1 with PT 8.51.
Firstly I encountered problems running grant.sql ,it showed that PSSTATUS, PSOPRDEFN, PSACCESSPRFL TABLE OR VIEW does not exist.(So i tried following "Using Data Mover, You can login in to the SYSTEM schema and EXPORT those tables (Using the 'EXPORT <RECORD>' instead of 'EXPORT *'). Then using Data Mover again, login to the SYSADM schema and IMPORT the data. NOTE: You will need two DMS scripts, one to export and one to import, as well as making sure you're using the right USER/PW combination" solution).
Secondly I can login to DATA MOVER using three combination of id's and psswds i.e system/oracle(original bootstrap mode id's) and PS/PS, SYSADM/SYSADM but the last two combinations also shows as bootstrap mode inside datamover.
Lastly i triesd to import the three tables from one schema to other by logging in to system/oracle as all three tables are present under this schema but when i run,
set log C:\REF0001_EXPORT.log;
set output C:\REF0001.dat;
export PSSTATUS;
It shows the below error in logs,
Invalid command in boot strap mode
Ended: Tue Jun 14 00:37:12 2011
Unsuccessful completion
Pls help!

Thanks Nicolas,
So now what do you advise me to do...Do i need to start a fresh from scratch??..I beleive half of the tables present r under SYSADM and rest half under SYSTEM..So is there anything which i can do to import those??

[Solved] Numlock on boot (with systemd)

I recently upgraded to systemd. Everything is working extremely well. Except that I am given to understand that systemd no doesn't use /etc/inittab, where I had added the command from the wiki that turns on numlock on each of the virtual terminals. Unfortunately the wiki doesn't have any method to do it with systemd, only the /etc/inittab method.
This means that for a brief while, I had to live with the onerous task of hitting numlock manually. It was one of the worst weeks of my life. My index finger is still raw.
I have come up with a somewhat-solution. I just edited /etc/bash.bashrc and added
if [[ $( tty ) == /dev/tty? ]]; then
setleds +num
fi
to it, which works quite well. It turns on num lock once you're logged in as long as you are on a virtual terminal and not on one of the /dev/pts/* ones.
My question is, is there a better place to do this? Maybe some way to get it to turn on before you're even logged in? It doesn't really matter all that much, I'm just trying to make things "correct."
Last edited by scott_fakename (2012-09-06 20:29:46)

DSpider wrote:
https://wiki.archlinux.org/index.php/Ac … _on_Bootup
Set it in Xorg for whatever DE/WM you use and it will be active when you switch to a tty.
Don't forget to mark it as solved.
Yes, I tried that. I use lxdm and xfce, and setting the numlock=1 option in /etc/lxdm/lxdm.conf turns on numlock in xfce and the light in the consoles, but not the actual numlock itself in consoles.
And Kejpi thanks, I was doing that but when I upgraded to systemd it said those would be taken out "at some point" so I was trying to find out if a permanent solution existed yet. Apparently it does not exist yet, which is fine, I was just curious. So I guess I'll stick with the bashrc method for now.
Thanks for the replies.
--Scott

Brightness buttons not working and glitchy with simple commands

I bought a macbook pro 6 months ago. The computer is in mint condition, but has been glitchy, especially with signing in or the simple command of putting in a login password. In order to login you must shut down the computer because the screen is frozen after a logout previously. On top of that the brightness keys show they work but the brightness doesn't change. This computer has no scratches, never been dropeed, has barely been used, and should be working perfectly. Any ideas?

You have a 1 year warranty, I would take it in and have them deal with it (you paid for it), they might find something that will really bite you further down the road. USE your warranty, they are usually quite good at customer support.

Help! My original and very expensive CS6 has crashed with a simple command of selection-.

Help! My original and very expensive CS6 has crashed with a simple command of selection….

Thanks JJMack,
unfortunately I'm Italian and I do not speak English, and most importantly
I'm not a genius in the world of computing. In fact, if you talks to me
about preferences folder User id and I do not even know where to look ....
why has crashed in your opinion? I was creating patterns with low
resolutions. It can if you have the time and patience to help me understand
where look the folder and what to do in a primitive language so that an
incompetent like me can understand.
Thank you in advance.
greetings
Vittorio.
Il 06/apr/2014 15:19 "JJMack" <[email protected]> ha scritto:
Re: Help! My original and very expensive CS6 has crashed with a
simple command of selection.... created by JJMack<http://forums.adobe.com/people/JJMack>in *Photoshop
General Discussion* - View the full discussion<http://forums.adobe.com/message/6275517#6275517

Solaris 10 Update 10 (8/11) panics on boot under virtualbox and vmware

I am trying to install S10 U10 in a VM and have downloaded the DVD ISO image. I verified the MD5 checksum then tried to boot the installation from the DVD image but it panics on boot under both Virtualbox (latest version) and VMware workstation (latest version). How can I get this working?
Thanks
Andrew.
Edited by: 884653 on 19-Sep-2011 13:18

I had no problems installing Solaris 10 U9 x86 from an ISO download. The ISO download of Solaris 10 U10 x86 I got recently is another matter. When I try to install U10 with VMware Player or Virtualbox on a Dell PC with Windows 7 the install process hangs. I can't tell if this behavior is a panic or not, but it is a problem. On both VMware Player and Virtualbox the install process gets to the following screen and never does anything else:
SunOS Release 5.10 Version Generic_147441-01 64-bit
Copyright (c) 1983, 2011, Oracle and/or its affiliates. All rights reserved.
Any ideas? Thanks.

Batch changes don't function. Apt 3.2. From simple commands like "rotate clockwise" to "lift and stamp" the command only affects to the last image in the group selected. Similar problems? Solutions?

Batch changes don't function. From simple commands like "rotate clockwise" to "lift and stamp" the command only affects to the last image in the group selected. Similar problems? Solutions? Suggestions? Thanks, Bruce

Glad it helped. Everyone has been bitten by this at one time or another. Fortunately you only get bit once
Well if you haven't already read it two or three times the User manual is really very good. Apple really did a nice job with it. The on-line version is nice because it is searchable.
After that the Apple Pro Training Series book Aperture 3 is very useful. Its basically split into three parts. The library, adjustments and sharing. In addition if you get all the way through it you can take the Certification test and become, well certified.
Rwboyer who posts here on and off has a good site and a couple of ebooks that I found useful.
good luck

Will latest mac mini boot under Tiger?

I have an older G4 tower set up for use as a graphic arts workstation. As far as the software set-up (both OS (Tiger) and applications (CS2)) goes, it's perfect for what I use it for and I've worked out all the bugs. The computer itself it's getting a little long in the tooth, however, and I'd like to replace that. I'm looking at the latest Mac mini. The ideal solution would be to simply clone the internal start-up drive to an external drive, then restore the clone onto the mini. The latest Mac mini comes (obviously) with Snow Leopard. Does anyone know if it requires Snow Leopard or whether it's possible to boot it with Tiger ?

Not sure how to "quote" into grey boxes but as far as this goes...
Place a > before the text you want to quote.
My MacBook Pro has an Intel processor and runs Tiger.
That Tiger OS is designed for Intel Macs. Tiger installs are PowerPC only unless created from the disks which shipped with an Intel Mac, in which case it will only install onto that specific Mac model, or the Mac OS X Server 10.4 Universal Binary.
If I reformat Mac mini's hard drive it wouldn't have any OS installed. I have unkeyed Tiger disks, so should then be able to install Tiger on reformatted mini drive, correct?
No.
is there anything in the hardware architecture of the latest Mac mini which will prevent it from booting under Tiger?
Yes. Pretty much the entire architecture prevents it. Give up.
(53112)

Simple command to refresh data of all declared internal tables & variables!

Hi All,
I have declared many internal tables & variables in my program in declaration include.
Is there any Simple command which can refresh data of all declared internal tables & variables!
Instead of clearing & refreshing each i.table & variable i want to use single command to refresh data.
Is this possible!
Thanks in advance.
Thanks,
Deep.

CLEAR: lv_field1,
             lv_field2,
             lt_itab1[],
             lt_itab2[].
ONE simple command (CLEAR), can initialize fields, workareas, field-symbols,.... and internal tables.

[SOLVED] cups not started under systemd

Under systemd cupsd is not started automatically; but manually run
systemctl start cups
start it. In trying to figure out what happens I see that the install section of cups.service contains the following:
WantedBy=printer.target
But nothing seems to depends on printer target. I have just changed that by
WantedBy=multi-user.target
and everything works. Anyway, either there is a bug that I have corrected, either I have missed something. The original printer.target seems to have nothing that depends on it. Could someone tell me what's the proper way to configure printing under systemd.
Last edited by olive (2012-12-14 06:51:27)

Now cups start properly and everything is fine. I do not know if is due to an upgrade or so, but I put solved now. Thanks for the explanation of the printer target.

Command on boot - systemd tmpfile

Hi.
I have a laptop with ATI/Intel hybrid graphics. I am never using the ATI graphics when in Linux, and i simply want to turn the ATI card off on boot for battery-life. I used to do this in Ubuntu, simply by putting the command:
echo OFF > /sys/kernel/debug/vgaswitcheroo/switch
in /etc/rc.local
Now on Arch i have to go around making a tmpfile for systemd (by searching around it seems this is the "correct" way of doing this kind of stuff on boot), that will perform this action for me. However i find the syntax very difficult to understand, so if someone could help i would appreciate it. The tmpfile i have created so far (which ofcourse isnt working), is
/etc/tmpfiles.d/fix_graphics.conf
and contains the line
w /sys/kernel/debug/vgaswitcheroo/switch - - - - OFF
Could anyone give me a pointer to what could be wrong? And just for future reference: What would one do in order to run a general command or script (for doing stuff more complex than writing a line to a file etc.) as root on startup with systemd?
Thanks in Advance
Esben

Hi
According to the wiki you have to blacklist the radeon kernel module and then load it manually afterwards before executing the echo OFF >...
Also you've to enable debugfs in your fstab, all described in the wiki: https://wiki.archlinux.org/index.php/Hy … ons_So_Far
As for systemd I do this with a little script that is called by a custom service file, wiki: https://wiki.archlinux.org/index.php/Sy … vice_files
I've added some After=, Requires= and Before= "tags" to the service file to make sure it executes at a good time, (After/Requires=systemd-user-sessions.service, to not execute too early and cause a kernel panic and Before=xinit.service to start it before X is started)
Edit: Alternatively you could use acpi_call from AUR to switch off the DGP completely. However this requires some similar steps to execute on boot.
Last edited by rebootl (2012-12-19 14:20:10)

{Solved} Can not successfully boot into systemd

I will first start off by saying I have read though the following threads:
https://bbs.archlinux.org/viewtopic.php?id=151633
https://bbs.archlinux.org/viewtopic.php?id=151885
And have tried to follow the Systemd wiki. I started out by switching to the new initscripts configuration system. I have the current systemd package installed. I then installed systemd-sysvcompat which of course had me remove sysvinit. I then removed initscripts which moved my rc.conf to rc.conf.pacsave.
I had to manually create /etc/hostname, /etc/locale.conf, and /etc/vconsole.conf as (mentioned in a previous post) I could not do with any "ctl" commands as I had not yet booted into a systemd setup.
This also caused an issue with trying to set up the timezone as:
# timedatectl set-timezone America/New_York
resulted in:
Failed to issue method call: Launch helper exited with unknown return code 1
So at this point I decided to try a reboot and I get:
Welcome to emergency mode! Type "systemctl defualt" or ^D to enter default mode.
Type "journelctl -b" to view system logs. Type "systemctl reboot" to reboot.
Give root password for maintenance
(or typed Control-D to continue): [ 71.520184] radeon_cp: Failed to load firmware
"radeon/R300_cp.bin"
[ 71.520205] [drm:r100_cp_init] *ERROR* failed to load firmware!
[ 71.520224] radeon 0000:01:05.0: failed initializing CP (-2)
[ 71.520237] radeon 0000:01:05.0: Disabling GPU acceleration
I entered emergency mode which brought me to the root terminal. I first attempted to mv rc.conf.pacnew to rc.conf expecting there to possibly be an issue with the daemons I had running or attempted to run but after reboot I got the same thing. Not sure what is going on with the radeon firmware because it's obviously there.
So I re-imaged the hard drive and decided to try again. This time I installed systemd-sysvcompat (removing sysvinit) and DID NOT remove initscripts. I rebooted and got the exact same thing as above. So at this point I am at a loss of ideas with moving forward. Why is systemd-sysvcompat causing the failure of the radeon firmware to load? I use Arch on my laptop, desktop as well as my server. Systemd is coming so I am trying with the laptop first to convert to pure systemd. If I am just missing a step please let me know and I appologize for posting. But hopefully others have run into this issue and could possibly give me a nudge in the right direction. It could be I am reading the wiki wrong or just not interpreting it correctly.
Thanks in advance for your guidance and assistance,
All the best,
Ian
Last edited by ichase (2012-11-16 16:55:10)

hiciu,
Here is what you requested taken from emergency mode:
/etc/mkinitcpio.conf
# vim:set ft=sh
# MODULES
# The following modules are loaded before any boot hooks are
# run. Advanced users may wish to specify all system modules
# in this array. For instance:
# MODULES="piix ide_disk reiserfs"
MODULES=""
# BINARIES
# This setting includes any additional binaries a given user may
# wish into the CPIO image. This is run last, so it may be used to
# override the actual binaries included by a given hook
# BINARIES are dependency parsed, so you may safely ignore libraries
BINARIES=""
# FILES
# This setting is similar to BINARIES above, however, files are added
# as-is and are not parsed in any way. This is useful for config files.
# Some users may wish to include modprobe.conf for custom module options
# like so:
# FILES="/etc/modprobe.d/modprobe.conf"
FILES=""
# HOOKS
# This is the most important setting in this file. The HOOKS control the
# modules and scripts added to the image, and what happens at boot time.
# Order is important, and it is recommended that you do not change the
# order in which HOOKS are added. Run 'mkinitcpio -H <hook name>' for
# help on a given hook.
# 'base' is _required_ unless you know precisely what you are doing.
# 'udev' is _required_ in order to automatically load modules
# 'filesystems' is _required_ unless you specify your fs modules in MODULES
# Examples:
## This setup specifies all modules in the MODULES setting above.
## No raid, lvm2, or encrypted root is needed.
# HOOKS="base"
## This setup will autodetect all modules for your system and should
## work as a sane default
# HOOKS="base udev autodetect pata scsi sata filesystems"
## This is identical to the above, except the old ide subsystem is
## used for IDE devices instead of the new pata subsystem.
# HOOKS="base udev autodetect ide scsi sata filesystems"
## This setup will generate a 'full' image which supports most systems.
## No autodetection is done.
# HOOKS="base udev pata scsi sata usb filesystems"
## This setup assembles a pata mdadm array with an encrypted root FS.
## Note: See 'mkinitcpio -H mdadm' for more information on raid devices.
# HOOKS="base udev pata mdadm encrypt filesystems"
## This setup loads an lvm2 volume group on a usb device.
# HOOKS="base udev usb lvm2 filesystems"
## NOTE: If you have /usr on a separate partition, you MUST include the
# usr, fsck and shutdown hooks.
HOOKS="base udev autodetect pata scsi sata filesystems usbinput fsck"
# COMPRESSION
# Use this to compress the initramfs image. By default, gzip compression
# is used. Use 'cat' to create an uncompressed image.
#COMPRESSION="gzip"
#COMPRESSION="bzip2"
#COMPRESSION="lzma"
#COMPRESSION="xz"
#COMPRESSION="lzop"
# COMPRESSION_OPTIONS
# Additional options for the compressor
#COMPRESSION_OPTIONS=""
The next items due to length are linked to pastebin;
bsdtar -t -f /boot/initramfs-linux.img
http://pastebin.com/cZEJHVHm
dmesg
http://pastebin.com/MUGs1e9H
journalctl -b
http://pastebin.com/xgbfVJP8
This is a lot to look at so do know that I am very appreciative of you and any one else looking at these files taking your time to help me with this issue.
All the best,
Ian
Edit: I also wanted to mention that I was able to get rid of the syslog-ng errors previously mentioned by performing:
systemctl enable syslog-ng.service
Then rebooting. Granted I was still put back in emergency mode but was able to fix one issue.
Last edited by ichase (2012-11-16 03:33:41)

[Solved]-[solved:indirectly]Chain loading services on boot in Systemd

I have what I believe is a simple issue, but am afraid that my ignorance of systemd services is preventing me from getting my desired result.
I want to boot into tty1-7 automatically and have them each launch independent chains of executions. Specifically, I'm trying to get tty1 to first log a user in and then to run a bash that I've written that loads fbv with an argument, and puts fbterm on top of it, and then load tmux...without me having to do anything.
Essentially I've gotten this sketched out for tty1:
[Service]
Type=oneshot
ExecStart=/home/<user_name>/fbterm-startbash ~/wallpaper.jpg
StandardInput=tty
StandardOutput=tty
TTYPath=/dev/tty1
I read through https://wiki.archlinux.org/index.php/Au … al_console and applied that, but there's no enabling of services there. Is there something simple that I'm missing about adding extra "ExecStart"s in the same service and how to time them?
(I put this in Apps and DE, because a terminal + tmux is my environment, let me know if that is incorrect.)
Last edited by SquidGuy (2013-06-08 14:08:56)

I'm not sure I can address your question directly, but I have an alternate - likely much simpler - solution.
Set up autologin to any of the ttys you want it on (see autologin in wiki if needed).
Then, in your shell profile (e.g, ~/.bash_profile), add a section with conditionals that check $XDG_VTNR:
if [[ $XDG_VTNR -eq 1 ]]; then
# stuff for tty1 here
elif [[ $XDG_VTNR -eq 2 ]]; then
# stuff for tty2 here
fi
EDIT: you may also want all that in a `if [[ $TERM == "linux" ]]` block.
Last edited by Trilby (2013-06-08 13:05:23)

Adding simple commands to boot under systemd

Similar Messages

Maybe you are looking for