breaks if system partitions are not reformatted
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
ubiquity (Ubuntu) |
Fix Released
|
High
|
Colin Watson |
Bug Description
Installing RC onto a serial ATA drive to /dev/sda10 mounted as /target. It seemed to have copied all files and was about to install the packages.
Traceback (most recent call last):
File "/usr/bin/
install(
File "/usr/bin/
ret = wizard.run()
File "/usr/lib/
self.
File "/usr/lib/
self.
File "/usr/lib/
self.
File "/usr/lib/
raise RuntimeError, "Install failed with exit code %s" % ret
RuntimeError: Install failed with exit code 1
Dennis Kaarsemaker (dennis) wrote : Re: [Bug 47046] Ubiquity crash | #1 |
Mark Constable (markc) wrote : Re: Ubiquity crash | #2 |
Unfortunately that livecd session is no longer available but I will know, next time, what log files to attach.
However, I think (95%) I know what may have caused this problem. The partition I was installing onto had (or may have had) a previous amd64 test installation (probably Gentoo) and when Ubiquity tried to chroot into /target to complete the installation it could have crashed if any x86_64 binaries or libs were called. When I checked the partition after this failed attempt there was a lib -> lib64 link in / and I did use this machine for testing various 64 bit installs over the last year. When I reformatted that partition and tried again the Ubiquity installation succeeded and I could boot on that particular partition.
So perhaps there is not a problem with Ubiquity and this was simply operator error. However, maybe a check for unsuccessful execution when chroot'ing could provide a warning.
Colin Watson (cjwatson) wrote : | #3 |
I should either enforce reformatting of /, /usr, and other system partitions, or I should put up a warning in the event that they are not reformatted.
Changed in ubiquity: | |
status: | Unconfirmed → Confirmed |
Mark Constable (markc) wrote : | #4 |
If you enforce reformatting of system partitions then please be careful. In my case above I was using /dev/sda10 only as a test install. There were still usable and wanted "real" 64bit system partitions on /dev/sda5 thru /dev/sda8 (sda9 is my swap on this drive).
My USB wireless card would not work at all with the RC amd64 installation so I was installing the regular i386 RC onto a spare partition on the same hardware to see if that would work with my wireless device (and it did, intermittently).
Colin Watson (cjwatson) wrote : | #5 |
Naturally we'd only force reformatting of partitions that you actually had mounted on /, /usr, or whatever, not random other ones lying around on the disk.
I think I prefer d-i's approach though - put up a warning and let the user decide whether to go ahead anyway.
Cristian Nastase (cnastase) wrote : | #6 |
The system breaks even the partitions ARE formatted. I got the same error ... and I forced formating the swap and / partions. Tha latter I did it as XFS. In fact it took me 1 hour to try to install. Qtparted was a complete mess. Kept trying to install on the ntfs partion. Could not find why the swap and / entries were missing for the life of me. So I used fdisk and mkfs /dev/hdaX -t xfs. Then the installer went ok till this bug. How do I attach the syslog and partman. I forgot to mention that I am talking about the x86_64 version.
Andre Landwehr (andrel) wrote : | #7 |
I have the same problem here, Dapper Drake final for AMD 64. I tried to install into fresh partitions, but use my old 32 Bit XFS /home without formatting it. Ubiquity crashed 100% reproducible. Now I install without using my old home and it seems to work. I'll try to mount the stuff by hand later when I have the system up and running.
Colin Watson (cjwatson) wrote : | #8 |
Cristian, you're encountering a different problem. See "Crash if /boot is on XFS" in https:/
Cristian Nastase (cnastase) wrote : Re: [Bug 47046] Re: breaks if system partitions are not reformatted | #9 |
Ha ... I actually fixed it the way is explained without knowing it ... I
found some old stuff on goggle xfs/ubuntu related where somebody was
mentioning that ubuntu needs a /boot as ext3 ... I did it and it worked
perfectly. That was yesterday ... Anyway, I will post later on today my
solution for reference (for it needs additional fix for qtparted). All
in all I have to take bow for kubuntu/community ...
Colin Watson wrote:
>Cristian, you're encountering a different problem. See "Crash if /boot
>is on XFS" in
>https:/
>you may well be having that problem too.
>
>
-------
Wessel van Norel (delgurth) wrote : | #10 |
You say you should reformat the partitions. But my experience with this bug was the following. If I created the mountpoints myself (with amd64 flight 6), I would get this exception. Both mountpoints where set to be formatted. If I told ubunto to just create the mountpoints by itself, nothing was wrong. But I don't want to use the entire disk for ubuntu, so that's not an option for me.
I hope you guys can work this problem out.
Wessel van Norel (delgurth) wrote : | #11 |
Oh, I should read better before commenting. I guess I've also experienced the XFS /boot problem.
I'll try it with a non XFS /boot partition, and I assume that will resolve my problem.
Colin Watson (cjwatson) wrote : | #12 |
Fixed in Edgy, to be released in October 2006. I've also included this fix in the list of changes scheduled to go into the first point release of Dapper, which will probably be in August 2006. Since there are a lot of people subscribed to this bug due to duplicates, I'd appreciate it if any queries about this bug or complaints about the form of the fix were sent by private e-mail or filed as new bugs respectively.
ubiquity (1.1.0) edgy; urgency=low
* Use try/finally in install.py to ensure that PROGRESS STOP is always
called; ignore errors from PROGRESS STOP (e.g. because debconf has gone
away due to a previous error).
* Support /UNIONFS (Knoppix) as well as /rofs as the read-only filesystem
source.
* Don't require entering a full name (see
http://
* GTK frontend:
- Return None from location_from_point if point is None (thanks to Chris
Moore for the idea; closes: Malone #48885).
* KDE frontend:
- Add slightly more paranoia in case getNearestCity returns None.
* Don't show hw-detect/
Malone #48524).
* Make it clear in the crash dialog that you should file a new bug, not
attach your details to an existing bug.
* Add passwd and adduser dependencies from user-setup.
* Take a copy of the supplied choice map in set_keyboard_
* Save tracebacks from install.py and slurp them back into the traceback
displayed if install.py exits non-zero.
* Convert to new Python policy (python-central). Install Python
modules/
public module path.
* Warn in the partitioner if you try to use XFS for /boot (closes: Malone
#47848).
* Warn in the partitioner if system filesystems are not reformatted
(closes: Malone #47046).
* Remove /usr/bin/espresso compatibility symlink.
* Use os._exit() rather than sys.exit() in the apt install progress child
process after forking, fixing a lot of strange "PROGRESS STOP" messages
in /var/log/
Malone #53298.
* Use exceptions to communicate all failures in install.py, so that we get
more useful tracebacks.
* Automatic update of included source packages: apt-setup 1:0.11ubuntu1,
base-installer 1.63ubuntu2, choose-mirror 2.04ubuntu1, clock-setup
0.10ubuntu1, debian-
1.17ubuntu1, hw-detect 1.38ubuntu1, kbd-chooser 1.29ubuntu1,
localechooser 1.20ubuntu3, partconf 1.16, partman-auto 54ubuntu1,
partman-base 90ubuntu1, partman-
partman-
partman-
partman-
silo-installer 1.04ubuntu1, tzsetup 1:0.10ubuntu1, user-setup
1.2ubuntu2, yaboot-installer 1.1.7ubuntu1.
* Handle move of prebaseconfig scripts to (variously) post-base-installer
or finish-install.
* Install search-path from debia...
Changed in ubiquity: | |
assignee: | nobody → kamion |
status: | Confirmed → Fix Released |
Stephane (mittiprovence) wrote : SV: [Bug 47046] Re: breaks if system partitions are not reformatted | #13 |
I just saw that.
Both / and /home need to be reformatted.
Thanks
Stephane
Colin Watson <email address hidden> skrev: ** Bug 54017 has been marked a duplicate of this bug
--
breaks if system partitions are not reformatted
https:/
Colin Watson (cjwatson) wrote : | #14 |
This will also be fixed in the first Dapper point release (which should hopefully dry up the flood of new reports of this problem!).
ubiquity (1.0.13) dapper-updates; urgency=low
* Don't require entering a full name (see
http://
* GTK frontend:
- Return None from location_from_point if point is None (thanks to Chris
Moore for the idea; closes: Malone #48885).
* KDE frontend:
- Add slightly more paranoia in case getNearestCity returns None.
* Don't show hw-detect/
Malone #48524).
* Make it clear in the crash dialog that you should file a new bug, not
attach your details to an existing bug.
* Take a copy of the supplied choice map in set_keyboard_
* Save tracebacks from install.py and slurp them back into the traceback
displayed if install.py exits non-zero.
* Warn in the partitioner if you try to use XFS for /boot (closes: Malone
#47848).
* Warn in the partitioner if system filesystems are not reformatted
(closes: Malone #47046).
* Use os._exit() rather than sys.exit() in the apt install progress child
process after forking, fixing a lot of strange "PROGRESS STOP" messages
in /var/log/
Malone #53298.
* Use 'replace' Unicode error handling when dealing with data coming from
debconf (closes: Malone #44595, #48732).
* Fix ubiquity/
ubiquity/
(closes: Malone #53258).
* Install more defence against gparted/qtparted crashing (closes: Malone
#47194, #48856).
* If gparted or qtparted crashes, put up a dialog allowing the user to try
again, return to autopartitioning, or quit the installer. This is a
pretty awful hack, but hopefully we won't be using gparted/qtparted for
all that much longer ...
* Fix crash in KDE frontend while translating widgets (closes: Malone
#53367).
* When an error occurs with the install progress bar up, only return to
autopartiti
* Turn DebconfInstallP
apt installation and the subprocess handles the translation of apt
status messages into debconf protocol messages, rather than the other
way round. The previous approach resulted in exceptions from apt ending
up in the wrong process.
* Silence apt errors while installing language packs, since we ignore them
anyway.
* Point apt at /target/
/var/
* Display grub-installer/
* Fix get_filesystems() not to assume ext3 in quite so many situations.
* Detect XFS for validation purposes even when it wasn't just formatted by
gparted/
* Fix all mentions of gparted in the KDE frontend.
* Leave existing newworld boot partitions alone when committing manual
partitioning changes.
* Always set mirror/country to the country selected on the timez...
Erich Graßmann (erich-grassmann) wrote : Bug 57730 - Installer crashed | #15 |
Hello Vassilis,
thanks for your answer.
It's correct, the partition for installing ubuntu wasn't formatted by the installer. But I thought it was a free ext3-partition, only formatted by SuSE Linux (9.x or 10.0). And now I remember there was a crashed Kubuntu-
I'll try it again (but not today).
Erich
(excuse my bad English, I'm better in German)
-----Ursprüngliche Nachricht-----
Von: Bug 47046 <email address hidden>
Gesendet: 26.08.06 12:55:22
An: <email address hidden>
Betreff: [Bug 47046] Re: breaks if system partitions are not reformatted
** Bug 57730 has been marked a duplicate of this bug
--
breaks if system partitions are not reformatted
https:/
_______
Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
http://
Rolf (rolf-kallenback) wrote : Re: [Bug 47046] Re: breaks if system partitions are not reformatted | #16 |
Please tell me how I unregister from getting these emails.
/Rolf
2006/8/2, Carthik Sharma <email address hidden>:
>
> ** Bug 54964 has been marked a duplicate of this bug
>
> --
> breaks if system partitions are not reformatted
> https:/
>
Vassilis Pandis (pandisv) wrote : | #17 |
The unsubscribe issues are known to the launchpad crew and they are working on them.... doesn't unsubscribing from bug 47046 help ?
MoMaT (momat) wrote : | #18 |
It is not possible to unsubscribe from bug 47046 (original bug) since one is subscribed only to its duplicate and is just "notified" about changes in the original one (it is not a proper subsription). To stop the e-mail flood one need to unsubsribe from the duplicate. See bug 48860 for more details.
Stephane (mittiprovence) wrote : SV: [Bug 47046] Re: breaks if system partitions are not reformatted | #19 |
How can I unsubscriibe from this mailing list..?
Regards,
Stephane
MoMaT <email address hidden> skrev: It is not possible to unsubscribe from bug 47046 (original bug) since
one is subscribed only to its duplicate and is just "notified" about
changes in the original one (it is not a proper subsription). To stop
the e-mail flood one need to unsubsribe from the duplicate. See bug
48860 for more details.
--
breaks if system partitions are not reformatted
https:/
Wolf Halton (saphil) wrote : Re: [Bug 47046] Re: breaks if system partitions are not reformatted | #20 |
Thanks!
Reformatting worked bonzer!
-Wolf
--- Vassilis Pandis <email address hidden> wrote:
> ** Bug 63413 has been marked a duplicate of this bug
>
> --
> breaks if system partitions are not reformatted
> https:/
>
--
Summer Special - Make Money on Your Phone Bill Arrowstars.com
Computer support network: http://
Eggs from Happy Chickens! Catwood Farms - 1960 Hightower Trail, Conyers GA 30012-1822 - 678-384-4930
_______
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://
Mårten Woxberg (maxmc) wrote : | #21 |
Isn't this bug fixed yet? I still get duplicate mails about it...
鸡毛 (liwf-net) wrote : | #22 |
me too...
On Wed, Mar 26, 2008 at 1:39 AM, Mårten Woxberg <email address hidden>
wrote:
> Isn't this bug fixed yet? I still get duplicate mails about it...
>
> --
> breaks if system partitions are not reformatted
> https:/
> You received this bug notification because you are a direct subscriber
> of a duplicate bug.
>
--
鸡毛/李文富
http://
Vincent Laisney (vincent-laisney) wrote : | #23 |
This bug is not fixed. I have had it today (Hardy-alternate downloaded yesterday).
On a Toshiba Satellite 1100.
I join the /var/log/syslog file I could get.
I hope it can help.
When I had reformatted the partitions. I had no more this bug (but another later!)
I hope it can help. Thanks a lot.
Vincent Laisney
鸡毛 (liwf-net) wrote : liwf wants to keep up with you on Twitter | #24 |
To find out more about Twitter, visit the link below:
http://
Thanks,
-The Twitter Team
About Twitter
Twitter is a unique approach to communication and networking based on the simple concept of status. What are you doing? What are your friends doing—right now? With Twitter, you may answer this question over SMS, IM, or the Web and the responses are shared between contacts.
This message was sent by a Twitter user who entered your email address. If you'd prefer not to receive emails when other people invite you to Twitter, click here:
http://
鸡毛 (liwf-net) wrote : | #25 |
To find out more about Twitter, visit the link below:
http://
Thanks,
-The Twitter Team
About Twitter
Twitter is a unique approach to communication and networking based on the simple concept of status. What are you doing? What are your friends doing—right now? With Twitter, you may answer this question over SMS, IM, or the Web and the responses are shared between contacts.
This message was sent by a Twitter user who entered your email address. If you'd prefer not to receive emails when other people invite you to Twitter, click here:
http://
Yan Maurakh (yan-maurakh) wrote : | #26 |
I tried to install kubuntu-netbook (karmic alpha 3) on my hp2140 mini note, and 'ubiquity' is still in error messages
Wolfgang Habermann (whabermann) wrote : Wolfgang Habermann/GUS/DE ist nicht im Büro | #27 |
Ich werde ab 23.07.2009 nicht im Büro sein. Ich kehre zurück am
12.08.2009.
Ich werde Ihre Nachricht nach meiner Rückkehr beantworten.
Please attach /var/log/ installer/ syslog and /var/log/partman to this
bug.