Bug 5381 - Safe mode boot hangs with "Transaction is destructive"
Summary: Safe mode boot hangs with "Transaction is destructive"
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: i586 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: Colin Guthrie
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks: 2120
  Show dependency treegraph
 
Reported: 2012-04-13 01:13 CEST by Morgan Leijström
Modified: 2012-04-28 16:08 CEST (History)
3 users (show)

See Also:
Source RPM:
CVE:
Status comment:


Attachments

Description Morgan Leijström 2012-04-13 01:13:29 CEST
This is a week old fresh network install of cauldron i586
It is working great, but now i wanted to try safe mode boot.

It seemed to work OK, but strangely it asked twice for encryption key.
Once while screen text was large, and once aftr it had switched to high res mode (sceen is 1600x1200)


Last two lines are:

Give root password for maintenance
(or type Control-D to continue): Failed to issue method call: Transaction is destructive.

And it only responded to ctrl-alt-del.



System uses a disk with a small /boot partition, a encrypted LVM with one partiton in it, and one non encrypted LVM with /home, swap in it.
All set up by mageia installer, but diskdrake have a slight problem with it, may be same problem, must file bug on that too...  may be related.
Comment 1 Morgan Leijström 2012-04-13 02:27:42 CEST
Seems i have multiple problems regarding partitions or something...
https://bugs.mageia.org/show_bug.cgi?id=5383
Comment 2 Bjarne Thomsen 2012-04-19 05:25:43 CEST
I have the same problem with a fresh instal in virtualbox of
Mageia-2-beta3-i586-DVD.iso
The install went well, but then I tried the safe mode twice and it stops with
Give root password for maintenance
(or type Control-D to continue): Failed to issue method call: Transaction is
destructive.

The installer chose the partitioning.

CC: (none) => bjarne.thomsen

Manuel Hiebel 2012-04-19 09:26:08 CEST

CC: (none) => mageia, pterjan

Comment 3 Colin Guthrie 2012-04-19 11:35:52 CEST
OK, I'll take a look. Thanks.

Status: NEW => ASSIGNED
Assignee: bugsquad => mageia

Comment 4 Bjarne Thomsen 2012-04-24 14:18:07 CEST
This same thing happens (beta3) also on a fit-pc2 when selecting "safe mode".
Only Ctl-Alt-Del can make it re-boot. I did not find any trace of the failed boot in messages. The standard boot mode works. I am not sure what to try next.
As for the hardware, here is output from lspcidrake:
rt3090sta       : Ralink corp.|RT3090 Wireless 802.11n 1T/1R PCIe [NETWORK_OTHER]
r8169           : Realtek Semiconductor Co., Ltd.|RTL8111/8168B PCI Express Gigabit Ethernet controller [NETWORK_ETHERNET] (rev: 02)
pata_sch        : Intel Corporation|System Controller Hub (SCH Poulsbo) IDE Controller [STORAGE_IDE] (rev: 07)
lpc_sch         : Intel Corporation|System Controller Hub (SCH Poulsbo) LPC Bridge [BRIDGE_ISA] (rev: 07)
sdhci_pci       : Intel Corporation|System Controller Hub (SCH Poulsbo) SDIO Controller #2 [SYSTEM_SDHCI] (rev: 07)
sdhci_pci       : Intel Corporation|System Controller Hub (SCH Poulsbo) SDIO Controller #1 [SYSTEM_SDHCI] (rev: 07)
ehci_hcd        : Intel Corporation|System Controller Hub (SCH Poulsbo) USB EHCI #1 [SERIAL_USB] (rev: 07)
uhci_hcd        : Intel Corporation|System Controller Hub (SCH Poulsbo) USB UHCI #3 [SERIAL_USB] (rev: 07)
uhci_hcd        : Intel Corporation|System Controller Hub (SCH Poulsbo) USB UHCI #2 [SERIAL_USB] (rev: 07)
uhci_hcd        : Intel Corporation|System Controller Hub (SCH Poulsbo) USB UHCI #1 [SERIAL_USB] (rev: 07)
shpchp          : Intel Corporation|System Controller Hub (SCH Poulsbo) PCI Express Port 2 [BRIDGE_PCI] (rev: 07)
shpchp          : Intel Corporation|System Controller Hub (SCH Poulsbo) PCI Express Port 1 [BRIDGE_PCI] (rev: 07)
snd_hda_intel   : Intel Corporation|System Controller Hub (SCH Poulsbo) HD Audio Controller (rev: 07)
Card:Intel Poulsbo US15W (GMA500): Intel Corporation|System Controller Hub (SCH Poulsbo) Graphics Controller [DISPLAY_VGA] (rev: 07)
unknown         : Intel Corporation|System Controller Hub (SCH Poulsbo) [BRIDGE_HOST] (rev: 07)
hub             : Linux 3.3.3-desktop586-1.mga2 ehci_hcd|EHCI Host Controller [Hub|Unused|Full speed (or root) hub]
hub             : Linux 3.3.3-desktop586-1.mga2 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub]
lirc_igorplugusb: Cesko|AVR309:USB to UART protocol converter (simple)
hub             : Linux 3.3.3-desktop586-1.mga2 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub]
usbhid          : Logitech|Logitech Illuminated Keyboard [Human Interface Device|Boot Interface Subclass|Keyboard]
hub             : Linux 3.3.3-desktop586-1.mga2 uhci_hcd|UHCI Host Controller [Hub|Unused|Full speed (or root) hub]
usbhid          : Logitech|USB Optical Mouse [Human Interface Device|Boot Interface Subclass|Mouse]
Comment 5 Colin Guthrie 2012-04-26 00:03:11 CEST
I can reproduce, but I'm not 100% sure what transaction it's talking about. I'll try and get some info.

Blocks: (none) => 2120

Comment 6 Colin Guthrie 2012-04-26 01:42:38 CEST
OK, so there are two problems here...

It seems that rescue.service calls sulogin which exits with code 15. It then attempts to startup default.target. It is this transaction that is ultimately destructive.

The first step is to ensure that sulogin works. Our version is horribly out of date so first step would be to try a newer one and see if it works better.
Comment 7 Colin Guthrie 2012-04-26 11:09:01 CEST
I've noticed a problem related to the fact that we were running sulogin rather than sushell as we should have been but I'm not confident it's the actual problem here. Will test some more later.

It also seems that under sysvinit, we would not prompt for the root password, just run a shell directly.

I'm not sure if this is intentional or not, but if all else fails I'll just do that.
Comment 8 Colin Guthrie 2012-04-26 23:48:20 CEST
OK, so I found the problem in our systemd package. We were not defining a macro properly that meant some mageia specific stuff inside m4 files that made up some of the default units was not run. Correting that means we do the same thing sysvinit does regarding the emergency login and all works nicely. Please feel free to reopen if you still have issues, but it worked fine on my test install where I was able to reproduce the issue before.

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

Comment 9 Morgan Leijström 2012-04-28 13:10:51 CEST
On my T42

Good: now it sucessfully get to a prompt.

However, it still asks two times for encryption password.
Maybe a not related issue, should i file another bug?
Comment 10 Morgan Leijström 2012-04-28 16:08:06 CEST
Password twice: Bug 5657

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