Watch out for todays upgrade......

[Fintan]

After upgrading today (with synaptic) I can't boot into my Feisty. I get to a bash telling me that apt-get is not installed. To install type apt-get install apt, which of course doesn't work because apt-get isn't installed ???
startx doesn't work either because it is not installed. :P
It sends me to the login screen which will not load my kde. Has the upgrade uninstalled my kde amongst other things??
Obviously recovery doesn't work either.

Any ideas??

[dibl]

Yikes --- that is ugly! I just ran Adept upgrade and update, then saw your post. I guess I should be afraid to re-start the xserver at this point .....

I did notice there was an Adept update on the list.

[michellembrodeur]

Yikes --- that is ugly! I just ran Adept upgrade and update, then saw your post. I guess I should be afraid to re-start the xserver at this point .....

I did notice there was an Adept update on the list.


I did the update and rebooted everything is fine. But at the end of the update it said a "NEW VERSION OF KUBUNTU IS READY" click NEXT to install. There is no next key. So how do I tell what version I have installed. There was a lot of updates today.

[dibl]

Whew -- that was a sweaty 3 minutes, but my system survived a reboot and all seems fine.

Yes, Michelle, I do remember seeing that message, and wondering "what the heck -- what "NEXT" button?". I guess a programmer was a little hung over or something ....

:-X

uname -r still returns "2.6.20-13-generic", so that doesn't look an upgrade to me ....

[Fintan]

Lucky you guys, poor me >:(

I used synaptic and didn't see that message or a "next" button for that matter.

Anyway I tried again and wrote some of the stuff the screen showed me:
bash: no job control in this shell
bash: groups: command not found
bash: lesspipe: command not found
bash: dircolors: command not found
etc, etc.

Well if anyone has an idea I'd really aprec. it

[dibl]

Fintan, I don't know enough Linux/unix to be of much help, but one notion that occurred to me is the possibility that your user configuration is all screwed up, but maybe not the main Linux system? Would it possible to login as the super user and make a new user, and see if he can do things that your old user can't do?

Just throwing out guesses here -- it looks pretty dismal ....

:'(

[Fintan]

HEH, heh, I tried that but forgot to mention what happens when doing startx in recovery mode:
startx is not installed, you can install it by typing apt-get install xinit ::)

I will try and login as super user from the login screen. I have never done that with kubuntu.
Is it user: sudo, password:xxxx ?

I don't think my user is corupted, I wouldn't know why it should be, but provided I can login I'll give that a shot.

[dibl]

Did you try aptitude? I think it is apt-get's big brother -- command structure all the same.

[Fintan]

No, but how can I without apt?

[dibl]

try "sudo aptitude autoclean"

then "sudo aptitude update"

then "sudo aptitude install"

then (if you're still upright) "sudo aptitude install apt-get"

or, get really crazy and "sudo aptitude dist-upgrade"

I'll pray ...

:P

[dibl]

Adept Notifier is suddenly telling me there are more updates. It must be because I am in North America that it is behind Europe. I'm VERY afraid now .... :P

[penguin.ch]

I'm VERY afraid now .... :P


On my test system (07.04 beta 1), "now and here" (i.e. 04.04. 15.00 in Switzerland), apt-get works like a charm ...

[Fintan]

Okay I'll give that a try

[dibl]

Hmmm -- I got my courage together and ran the Adept Updater. It said there were 28 upgrades (40MB). But, upon completion of the installation process, there were unresolved dependencies, so no change. I think I will try Synaptic ... I'm too curious for my own good.... ;D


EDIT: Synaptic agrees with Adept -- there are unresolved dependencies. No upgrades happening here, near term.

[Fintan]

Okay, in recovery mode: sudo aptitude autoclean gave me:
sh: groups: comand not found
bash: sudo: comand not found

I then tried: aptitude autoclean which gave me: aptitude not found, to install type apt-get install aptitude, blah, blahdiblah.

Anyone?? Please!

[Praxxus]

Fintan, it could be as simple as your PATH settings being screwed up. What happens when you
echo $PATH from the command line? If you get a blank line, try
export PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games" and then see if bash can find apt, etc.

[Fintan]

Okay I'll give that a roll :)

[MepisReign]

After the upgrade my system did BOOT with no problems except for some broken dependencies on KDM and JAVA plugin, the adept notifier remained on the taskbar. After that I typed on the Konsole "sudo --configure -a" in order to install the pending packages (that ones that actually were downloaded without problems).
So far the system is booting fine, I will wait for that issues to be solved before take any "drastic" :P actions ;)

[Fintan]

Fintan, it could be as simple as your PATH settings being screwed up. What happens when you
echo $PATH from the command line? If you get a blank line, try
export PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games" and then see if bash can find apt, etc.


Okay did that, tried to reboot and got two messages that told me something about a bad usb cable. I didn't have any devices pluged in so I gave up and did a clean install. I am downloading the updtaes/-grades now and hope for the best.

I could have delt with broken depencies but not this.

Any way thank you all for your support.

Cheers
Fintan

[MepisReign]

Humm, sorry to hear that man, hopefully the clean install will solve the issues that the system may have, in any case those are the problems that as Beta users we are exposed to suffer :(

Take Care and Good Luck

[Fintan]

Yeah I know, and I'm not gripping. The clean install is still stable and I am going to stay away from updates/-grades for a while, although I don't know if I can control myself ;)

It would be a nice improvement to see a real system rescue in the alt cd instead of the one offered. Maybe something like the one suse and some other distros have.

Again thanks for the help and support

Cheers
Fintan

[dibl]

Bummer, Fintan -- sorry.

Here's a little p.s., for whatever it is worth. I'm completely current as of 1600 hours, 4 APR, Eastern U.S. time -- neither Adept Notifier nor the Adept Manager shows anything "pending" after I run updates and upgrades and dist-upgrades.

HOWEVER, upon reviewing the packages listed in the Adept Manager, I see the linux-2.6.20-14-generic packages. I'm running 2.6.20-13-generic. Isn't that odd -- they are sitting there -- theoretically I could select them for installation, but the dist-upgrade command returns nothing available.

I think I'll sit tight with a system that is working really well, for a day or two -- I'm suspicious of the situation with the -14 packages.

[Detonate]

Saw this thread early this morning, and decided to wait before updating. Just ran update, upgrade, and dist-upgrade with aptitude, rebooted with no problems. But it did reboot to -13, -14 was not listed in the grub menu. 1630 hrs, mdt, usa.

[nejode]

Well, just for the record, I upgraded 88 packages (158 mb) as of 15:00 local (Venezuela, GMT-4 hours), just read this post, rebooted to kernel -13 (as dibl said, -14 kernel is in the repos but there was no kernel upgrade) and found no problems whatsoever.

Usually, with a regular release, I'll wait for a couple of days before upgrading anything that has to do with kernels, xorg, apt, etc, and that saved me from the xorg-xserver upgrade chaos and the KDE 3.5.4 mess a few months ago... but with these beta's, upgrades come out so quickly that if you wait 2 days, maybe you'll have to deal with a 200+ mb download!

[dibl]

Fintan, I reproduced your failure! But, you did not say that you added or modified disk partitions on another drive, so maybe I got the same result via a different causal factor.

I decided to give the 64-bit Feisty Beta a trial run, since I have a Core 2 Duo with the EM64T architecture. To make a place for it, I removed a large partition on a hard drive and split it into two pieces, one for / and one for /home. The experiment went great -- of course a new Grub menu was written and the new OS promoted itself to the top of the list, etc. etc. The new OS runs great, but a benchmark test that I use revealed that it doesn't run my app any faster than the -generic OS.

However, when I rebooted and picked my previous OS from the bottom of the Grub list, it failed to boot in EXACTLY the manner that you reported. FSCK went perfectly, then it reported that it would not mount "UUID-e7960f07-be4a ......... -- no such device"! Then came the same errors you saw -- no apt-get, no job control, no groups, no bash commands ... blah blah blah, just as you said. Left me at the root login, with barely the ability to shut it down.

I worked on the problem a little while and figured out that the change of partitioning on the second drive resulted in a wrong fstab file on the first drive (the original OS). When the original OS was trying to mount the drives in /media, the fstab map was no longer true to the actual partitions out there, due to the change I had made. I recovered by editing the fstab file in the original Feisty installation and changed "UUID=xxxxxxx" to "'/dev/sdc2" and then I needed to add the additional partition that I had created as /dev/sdc3, and VOILA I was back in business.

So, there is a way to cause the problem and a way to fix it, should it ever happen again.

;)

[shadowfx78]

in synaptic click on the edit menu and look for the fix broken packages option.

[penguin.ch]

changed "UUID=xxxxxxx" to "/dev/sdc2"


In additition to that, I'd recommend to replace all (so-called) UUIDs by the original devices nodes (in /etc/fstab as well as /boot/grub/menu.lst), at least to eliminate this well-known source of trouble ...

[Fintan]

@dibl Thanks and of course you are right. I did delete a partition at the END of my disk so I didn't think that would change anything with the partitions in front of it. Well I suppose it did. Stupid me for not changing fstab and grub like i did in edgy (as mentioned by unicorn).
I will do that know.

Thanks again.

Well now I have another problem. After reinstalling and rebooting twice without a problem I now get this message: hub 4-0:1,0: Cannot enable port2. Maybe the USB cable is bad?

I connected my scanner this morning because I wanted to test it. It worked fine before the updates.

I can access the feisty partition from edgy.

Any Ideas?

I'll move this to a different post.

[gregconquest]

I also got the "apt-get" is not installed. I haven't changed any partitions since this version of kubuntu was working. I can refresh my kubuntu image (backup from BootIt NG), have no problems, run "sudo apt-get update" click on the update icon, run through the update, and at reboot, I get the errors mentioned every time.

I also can't update successfully anymore. Some packages never complete no matter how many I select. I'm running through one now. I'll refresh my image, re-try the whole process, and edit this post with the notes . . .

Is there anything in particular I should be looking for? kernel version? apt-get update?

I did have a problem with fstab a few days ago as well. The offending partition was commented out -- perhaps before the install I'm refreshing now.

Greg

Update: my earliest image has some of the same problems. I'm going to re-install fresh.

[gregconquest]

This is no good. I have all the same problems even from a fresh install:

- I rearrange my partitions (to eliminate post-install partition change variable)
- reinstall kubuntu-7.04-beta-alternate-i386
(from April 1 -- do these CD images change
daily -- or will it stay the same till beta2?)
- grub installed to root (sda4), BootIt NG is in mbr
- reboot
[704ku001.img made]
- boot into 2.6.20-generic
- 49000 days since check -- fixed (1st sign of problems?)
- oem login
- sudo oem-config-prepare
- reboot
[704ku002.img made]
File System Check Failed (the erros are all the same from here)
A log is being saved in /var/log/fsck/checkfs if that location is writeable. Please repair the file system manually.
* A maintenance shell will now be started. CONTROL-D will terminate
bash: no job control in this shell
bash: groups: command not found
sh: groups: not found
("lesspipe" and "The" also not found)
The program apt-get is currently not installed. You can install it by typing: apt-get install apt
- I type EXIT and hit return
- boots into 1st login, my username and all accepted
- KDE launches, login
- adept_updater is waiting -- flashing an update.
- launch updater
install 8, upgrade 255 (incl. adept, adept-batch -common -installer -notifier -updater,
linux generic from 2.6.20.12.8 to 2.6.20.14.12)
- apply updates
- Bluetooth - failed to connect to the SDP server
- Could not commit changes - Adept Updater
Now I'm back to where I was before. No updates will complete.

Greg

[penguin.ch]

File System Check Failed


An apparently corrupted file system is what you have - and not an assumed problem with package management ... Therefore, please do as being told by your system and repair this file system manually :P

[gregconquest]

File System Check Failed

An apparently corrupted file system is what you have - and not an assumed problem with package management ... Therefore, please do as being told by your system and repair this file system manually :P


Well, this has gotten me to notice one thing. I was allowing the installer to mount my windows ntfs swap file! Thinking about it, I know that wouldn't be useful. I'm doing a fresh reinstall more carefully choosing and naming all mounted partitions.

Greg

UPDATE:
Mounting the partitions did help, but I still got the following errors:
- Bluetooth -- failed to connect to the SDP server (I'm installing without the bluetooth dongle from now)
- Could not commit changes -- Adept Updater. There was an error committing changes. Possibly there was a problem downloading . . . or the commit would break . . .

I'm thinking now that some of the problems were independent of my mounting quirks. It does appear that the first update is a problem. I'll do a piecemeal update now . . .

[gregconquest]

I refreshed back to the point where I had received no errors, and then launched updater.
I skipped fetching list of updates.
I successfully updated all packages except the following:
- akreagator
- amarok
- amarok-xine
- app-install data
- apport
- apport-qt
- avahi-autoipd
- avahi-daemon
- bluez-cups
- bluez-utils
- britty
- ttf-opensymbol
- ttf-thai-tlwg
After updating and applying all the other patches, I still had not received any errors.

A few other trials:
- selecting "ttf-opensymbol" alone resulted in "could not commit changes"
- selecting "akreagtor" alone resulted in "a new version is available" with the grayed out "next" button.
- selecting "amarok" and "amarok-qt" resulted in "a new version is available" with the grayed out "next" button.

I'm still testing the others. I have done updates of just a few without getting the errors above . . .

Greg

UPDATE: I kept getting the same errors by adding updates one-by-one. Finally, I refreshed again, launched the updater, told it to fetch new updates, then applied them all. I did not get the "could not commit" error, however I am still told that there is a new version available . . . click "next" . . . yet there is no "next" button to click.

[greenonion]

I got the same problem.
I noticed a message saying to hit ctrl-d to leave the maintenance shell and continue. I finished booting then, but I don't know what caused the problem. the system can't see my second hardrive though, and that was the source of the errors. Not sure how to do a disk check on it. Can anyone help me out?

Thanks

[penguin.ch]

http://www.die.net/doc/linux/man/man8/fsck.8.html

[greenonion]

The hard drive was seen fine, and loaded fine up until I installed xine-lib-1.1.4. The install loaded several different files, so I am not sure which one may be causing the problem. I just get the error no such drive when I try to mount it.

[mmcmcmc]

Hi

I've been having the same problem as mentioned above. The system boots in to a root session with the harddrive en read-only. The system sugested that I run fsck manually, which I've done some five times now, including running badblocks. Futhermore I downloaded the harddrive fitness test from IBM, which came out without any errors. These problems appear both in normal- and recovery-mode, in all the kernel versions that are installed.

Is reinstalling the only solution?