Bug 14028 - network install: urpmi error: ERROR: 'script' failed for systemd-216-1.mga5.x86_64:
Summary: network install: urpmi error: ERROR: 'script' failed for systemd-216-1.mga5...
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: Installer (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
Priority: release_blocker critical
Target Milestone: ---
Assignee: Colin Guthrie
QA Contact:
URL:
Whiteboard: after 5alpha2
Keywords:
Depends on:
Blocks: 14069
  Show dependency treegraph
 
Reported: 2014-09-01 16:35 CEST by Marja Van Waes
Modified: 2014-10-23 15:29 CEST (History)
6 users (show)

See Also:
Source RPM: systemd-216-1.mga5.x86_64
CVE:
Status comment:


Attachments
ddebug.log.xz (335.90 KB, application/x-xz)
2014-09-01 16:35 CEST, Marja Van Waes
Details
journalctl -a of normal + multi-user.target + rescue boots (386.37 KB, text/plain)
2014-09-03 11:40 CEST, Marja Van Waes
Details

Description Marja Van Waes 2014-09-01 16:35:07 CEST
Created attachment 5390 [details]
ddebug.log.xz

Yesterday, installing with yesterday's boot-nonfree.iso, there was an error message:

ERROR: 'script' failed for systemd-216-1.mga5.x86_64:

I can't attach a full report.bug.xz, because it was impossible to finish the install (it didn't manage to install an EFI-bootloader) and, of course, I couldn't find a USB key

Hope the attached and compressed ddebug.log is enough.

CC'ing leuhmanu, who hit this issue, too. Maybe he has better logs than me.
Comment 1 Marja Van Waes 2014-09-03 11:40:52 CEST
Created attachment 5405 [details]
journalctl -a of normal + multi-user.target + rescue boots

Just in case it is usefull at all (is it???): attaching "journalctl -a" output from the installed Mageia of 
* a normal boot at 10:53h CEST (which crashed), 
* a multi-user.target boot (impossible to log in) at 10:55h
* the rescue boot at 11:03h, which worked fine.
Comment 2 Georges Eckenschwiller 2014-09-03 11:52:40 CEST
I confirm the problem. I have the same message during the net-installation.
I also have:
script failed for cdrkit-1.1.11-4, cdrkit-genisoimage and samba-client.
By reboot, I have grub, plymouth screen, then a wait, then the computer stops.

CC: (none) => paiiou

Comment 3 Jüri Ivask 2014-09-04 06:39:54 CEST
I have a similar experience with network install (minimal install with X) using boot.iso dated September 3rd. "Script failed for systemd and samba-client"
Now the boot is very unreliable . Mostly it reaches DM but before I can login it dissapeares and computer shuts down. Sometimes it freezes after DM window dissapears and at about 20% cases DM window reappeares and I'm able to login.

CC: (none) => jyri2000

Comment 4 Georges Eckenschwiller 2014-09-06 13:17:36 CEST
I suggest to incrementing the severity, because it is not possible to restart correctly.

Severity: normal => critical

Comment 5 Marja Van Waes 2014-09-06 13:28:30 CEST
(In reply to Georges Eckenschwiller from comment #4)
> I suggest to incrementing the severity, because it is not possible to
> restart correctly.

Yeah, you're righ, this not being possible to boot normally into the new install makes it a release blocker, too.

CC'ing QA team leaders
Added "after 5alpha2" to the whiteboard, so it'll show up when you search for 5alpha2 release blockers.... it'll probably get fixed before 5beta1 QA testing starts, though ;-)

Whiteboard: (none) => after 5alpha2
Priority: Normal => release_blocker
CC: (none) => davidwhodgins, eeeemail, wilcal.int

Comment 6 claire robinson 2014-09-06 17:38:45 CEST
Thanks for adding us Marja. IIRC from dev ml there were some issues Colin was fixing with the new systemd. We can certainly test a network install when beta1 testing begins though in a few weeks.
Comment 7 William Kenney 2014-09-06 18:30:39 CEST
There are actually quite a few problems with network
installing using boot.iso or boot-non-free.iso.
Maybe this is a good time for me to go through
again and document them. One is that both boot.iso
and boot-non-free.iso do not work if you attempt
an ftp install where the ftp point requires an
ID & PW. Also if you attempt an http install
and the local repo, like mine, uses something
other then a common URL like:

http://mirrors.kernel.org/mageia/distrib/4/i586

Example, attempting to install from:

http://192.168.1.2:8080/~mageia/distrib/4/i586

Will not work. It used to work in M3 but M4 lost
that ability.
Comment 8 Manuel Hiebel 2014-09-06 20:46:24 CEST
Could we please stay on the original issue ?
claire robinson 2014-09-08 17:36:47 CEST

Blocks: (none) => 14069

Comment 9 Georges Eckenschwiller 2014-10-21 09:27:35 CEST
For me, the problem is solved.

Thanks
Comment 10 Colin Guthrie 2014-10-21 10:30:08 CEST
Cool. If someone else can confirm and close the bug that would be great! :)
Comment 11 Marja Van Waes 2014-10-22 09:23:24 CEST
(In reply to Colin Guthrie from comment #10)
> Cool. If someone else can confirm and close the bug that would be great! :)

Sorry, can't test yet because I have only WPA encryopted wlan where I am now (so impossible to connect when using boot.iso)

So I should create a private mirror (which was on my todo-list, anyway), or I should wait for the QA pre-beta1 traditional iso.

I'll try to go for both options and just use what I get first.

Btw, according to mirror.readme 
> The expected size of the mirror is around 700GB.

If that's really all, then I don't need to go out and buy a new ext. HD
Comment 12 Marja Van Waes 2014-10-23 15:29:05 CEST
(In reply to Colin Guthrie from comment #10)
> Cool. If someone else can confirm and close the bug that would be great! :)

The systemd script error is fixed in today's pre-5beta1 dual iso, so closing 

(In reply to Jüri Ivask from comment #3)
> I have a similar experience with network install (minimal install with X)
> using boot.iso dated September 3rd. "Script failed for systemd and
> samba-client"

That issue still exists 
If we don't have a separate report for it, then MrsB will create it

Status: NEW => RESOLVED
Resolution: (none) => FIXED

Comment 13 Marja Van Waes 2014-10-23 15:29:29 CEST
s/that/the samba-client/

Note You need to log in before you can comment on or make changes to this bug.