Bug 19554

Summary:	drakx sometimes gets a mismatch for the /home and /swap partitions on automatic partitioning when reinstalling the same mga version
Product:	Mageia	Reporter:	Kristoffer Grundström <lovaren>
Component:	Installer	Assignee:	Mageia tools maintainers <mageiatools>
Status:	RESOLVED WORKSFORME	QA Contact:
Severity:	normal
Priority:	Normal	CC:	lovaren, marja11, thierry.vignaud
Version:	Cauldron	Keywords:	NEEDINFO
Target Milestone:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Source RPM:	drakx-installer-stage2	CVE:
Status comment:
Attachments:	report.tar.gz

Description Kristoffer Grundström 2016-10-08 23:53:58 CEST

Description of problem: If I at the partitioning step in either Mageia 5 or Cauldron choose to let diskdrake automatically create partitions and then after the installation boot any kernel I end up to Emergency mode. This is due to not having working /home and /swap partitions. If I comment out the lines in /etc/fstab to not mount /home and /swap I get passed that error and that gives me a login screen. 

Version-Release number of selected component (if applicable): 16.105

How reproducible: Every now and then.

Steps to Reproduce:
1. Boot the full DVD iso or the small nonfree net-install.iso for Mageia 5.
2. Delete the current partitions, all but the /boot/EFI one.
3. Click to allocate new partitions automatically.
4. Install Cinnamon, MATE, KDE4 and GNOME.
5. After the installation is done, reboot your computer.
6. Boot the kernel.
7. During the kernel boot you'll see that /home and /swap fails to mount due to a missmatch of some sort.
8. Voila! Emergency mode is the only thing you can get to.

Comment 1 Kristoffer Grundström 2016-10-08 23:56:13 CEST

I can't reproduce it by hand so it just happens at some installations.
I get the same result if I choose to make my own partition table.

Comment 2 Marja Van Waes 2016-10-09 08:24:51 CEST

please attach 

      /root/drakx/report.bug.xz

from such an install

Keywords: (none) => NEEDINFO
CC: (none) => marja11
Assignee: bugsquad => mageiatools

Comment 3 Marja Van Waes 2016-10-09 08:27:57 CEST

and please attach /etc/fstab too :)

Comment 4 Kristoffer Grundström 2016-10-12 08:33:17 CEST

Created attachment 8521 [details]
report.tar.gz

CC: (none) => hamnisdude

Comment 5 Kristoffer Grundström 2016-10-12 08:35:39 CEST

I had to comment the /home and /swap paritions in order to get a working login window to be able to login to MATE, but here's the output from /etc/fstab:

# Entry for /dev/sda2 :
UUID=6c0d942c-1940-4533-a26a-2875729bf9e7 / ext4 noatime,acl 1 1
# Entry for /dev/sda1 :
UUID=4CD5-943E /boot/EFI vfat umask=0,iocharset=utf8 0 0
# Entry for /dev/sda4 :
#UUID=2dea2a2c-3f2e-40cf-b046-7000e8d65434 /home ext4 noatime,acl 1 2
none /proc proc defaults 0 0
# Entry for /dev/sda3 :
#UUID=0ed602b3-ce14-4d00-bdca-d2466643e4e7 swap swap defaults 0 0

Comment 6 Kristoffer Grundström 2016-10-14 04:49:23 CEST

Installing with the Mageia-Cauldron-netinstall-nonfree-x86_64.iso doesn't reproduce this problem.

I went from Ubuntu 16.04 to Mageia 6 (Cauldron).

The only time I can reproduce this as I see it is if I have Mageia 5 installed and try to install Mageia 5 again using the Mageia-5-x86_64-DVD.iso.

What more info do you need?

Thierry Vignaud 2016-10-14 10:58:18 CEST

CC: (none) => thierry.vignaud
Summary: drakx sometimes gets a missmatch for the /home and /swap partitions on automatic partitioning => drakx sometimes gets a missmatch for the /home and /swap partitions on automatic partitioning when reinstalling the same mga version

Thierry Vignaud 2016-10-14 10:58:28 CEST

Summary: drakx sometimes gets a missmatch for the /home and /swap partitions on automatic partitioning when reinstalling the same mga version => drakx sometimes gets a mismatch for the /home and /swap partitions on automatic partitioning when reinstalling the same mga version

Comment 7 Thierry Vignaud 2016-10-14 16:55:25 CEST

According to your logs, those as the right UUIDs after formatting, so it's not a matter of keeping old UUIDs.
We would need more details about what goes havroc.
Could you do a boot with those entries uncommented, then collect logs ("journalctl -b"  will gives you all the logs from the boot)

Comment 8 Kristoffer Grundström 2017-06-25 04:58:20 CEST

I think we can close this report since I at the latest installation of Cauldron didn't get that problem. I will however reopen this if I get the same problen with a fresh installation using the stable Mageia 6 iso.

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