Bug 414 - Bootloader error during installation, An error occurred, mkinitrd failed
Summary: Bootloader error during installation, An error occurred, mkinitrd failed
Status: RESOLVED OLD
Alias: None
Product: Mageia
Classification: Unclassified
Component: Installer (show other bugs)
Version: Cauldron
Hardware: i586 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2011-03-17 06:47 CET by Stephen Pettin
Modified: 2012-01-20 23:19 CET (History)
6 users (show)

See Also:
Source RPM: mkinitrd
CVE:
Status comment:


Attachments
Screenshot of log during install in VirtualBox showing mkinitrd failed. (203.86 KB, image/jpeg)
2011-06-05 20:33 CEST, Dave Hodgins
Details

Description Stephen Pettin 2011-03-17 06:47:03 CET
I want to install alpha2 inside Virtual Box 4.04 as guest and using Mandriva 2010.2 as host. Everything seems to be installing fine until I get to the installing bootloader screen, then it stops with the message below.

Bootloader error:

An error occurred
mkinitrd failed:
(mkinitrd -v -f /boot/initrd-2.6.37-desktop586-1.mga.img 2.6.37.3-desktop586-1.mga)

If I click the OK button, it loops back to the bootloader screen, I have to stop the process completely to get out of the loop. I also had this same issue with alpha1.

Not sure if this is a good idea at this stage for installing in VBbox.

My hardware I'm currently using is a three year old Acer Aspire One netbook with 1gb ram, 160gb hd, intel video card.

Thnx.


Reproducible: 

Steps to Reproduce:
Comment 1 Dave Hodgins 2011-03-17 08:14:47 CET
See https://bugs.mageia.org/show_bug.cgi?id=44

In alpha2, the problem is worse.  The user never gets the
option to select alt1 to use the sever kernel.

I have not yet figured out why none of the server kernels
will not work in VirtualBox on a single core host.  All
server kernels I've tried stop at "booting kernel"

The current kernel used by the mageia install iso is not
usable on a single core system, running under VirtualBox.

I have not been able to get the install from iso working
on real hardware, and at this point have had to give up
on mageia, till this is fixed.

I do still have a virtualbox version installed, which
works ok using the custom kernel my system spent 10 days
compiling, but see no further point in me trying to test
mageia.  My dvd drive can read, but anything it trys to
write cannot be read.  Until I replace the dvd burner,
I cannot further test the mageia instlaller, except to
say it doesn't work from an iso image on the hard drive,
or booting from a usb flash drive created using
unetbootin-linux-502.

After I replace my dvd burner, I will try magiea again,
but till then, there is no point in my testing it.

CC: (none) => davidwhodgins

Comment 2 Stephen Pettin 2011-03-25 11:07:53 CET
Sorry for a late reply. Thanks for the info/tips. I guess I'll wait till the first beta comes out and try it again. I'm also looking to get a tower system, since I'm using my current netbook as a main system. 

Thnx.
Comment 3 Stephen Pettin 2011-04-10 09:52:41 CEST
Well, I did get another system, a HP DX5150 and installed Mageia to the hard drive without issues. I did use the Beta 1 iso instead of the Alpha.

I haven't tried the beta in VirtualBox yet.

Just a thought!
Comment 4 Pascal Terjan 2011-04-26 00:10:13 CEST
mkinitrd gets killed if it takes more than 10 minutes, we can probably increase the timeout

CC: (none) => pterjan

Comment 5 Frédéric "LpSolit" Buclin 2011-06-05 12:38:05 CEST
Beta, RC and final installed without any problem in VirtualBox 4.0.8. Maybe is the problem fixed now?
Comment 6 Dave Hodgins 2011-06-05 20:33:38 CEST
Created attachment 523 [details]
Screenshot of log during install in VirtualBox showing mkinitrd failed.

Still a problem with Mageia 1.

The only way this can be fixed, is to provide (and have the installer use)
a kernel with CONFIG_HZ_100=y that does not require PAE (as VirtualBox does
not enable PAE by default.

While increasing the timeout for mkinitrd from 10 minutes to say 20, might
be enough to allow installation under VirtualBox, it still results in a
system that is so slow, it gives a really bad impression.

In my opinion, the best fix would be to add a kernel-lowend pacakage that is
configured the same as kernel-desktop, except for the HZ setting, and have
the installer use that kernel, during installation, and as the default when
installed under VirtualBox.
Comment 7 Dave Hodgins 2011-06-05 21:04:56 CEST
I ran the mkinitrd after booting into rescue mode, and it took 12min, 4sec.
Manuel Hiebel 2011-07-27 17:51:09 CEST

Source RPM: mkinitrd -v -f /boot/initrd-2.6.37-desktop586-1.mga.img 2.6.37.3-desktop586-1.mga => mkinitrd

Comment 8 Dave Hodgins 2011-07-27 21:39:00 CEST
While it is mkinitrd that is failing, mkinitrd is not the cause of the problem.

The problem is that the kernel used while running the installer is using
CONFIG_HZ_1000 instead of CONFIG_HZ_100. For background, see
http://www.virtualbox.org/manual/ch12.html#idp12796768

The kernel used when running the installer shoud use CONFIG_HZ_100, but
should not require PAE.

The kernel that gets installed can still use CONFIG_HZ_1000, if appropriate
(i.e. not running under VirtualBox and not on a slow cpu), but the installer
itself must be able to run on slow systems, and under VirtualBox.

A workaround would be to modify line 87 of /usr/lib/libDrakX/run_program.pm
to increase the timeout from 10 minutes to 15 or 20, but taking 2 hours or
more to do a minimal install (just selecting config and console tools), is not
a pleasant experience.
Ales Cerny 2011-09-09 16:32:16 CEST

CC: (none) => alda.cerny

AL13N 2011-10-30 13:30:50 CET

CC: (none) => alien, tmb

Comment 9 Dave Hodgins 2011-10-31 01:05:36 CET
Stephen, can you confirm that this is fixed by adding
divider=10 as a kernel option for the installer solves this issue?

If so, then this bug should be closed as a duplicate of bug 44.
Manuel Hiebel 2011-11-11 01:02:29 CET

Keywords: (none) => NEEDINFO

Comment 10 Marja Van Waes 2012-01-03 07:49:53 CET
@ Stephen,

Please reply to the question above within two weeks from now, to avoid this bug being closed as OLD.

CC: (none) => marja11

Comment 11 Marja Van Waes 2012-01-20 23:19:16 CET
Since there are insufficient details provided in this report for us to investigate the issue further, and we have not received feedback to the information we have requested above, we will assume the problem was not reproducible, or has been fixed in one of the updates we have released for the reporter's distribution.

Users who have experienced this problem are encouraged to upgrade to the latest update of their distribution, and if this issue turns out to still be reproducible in the latest update, please reopen this bug with additional information.

Closing as OLD.

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


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