I dared to update my Linux system to openSUSE 12.1 yesterday. The upgrade process went almost cleanly.
But booting does not succeed after this update at the moment.
The desktop kernel 3.1 can be started and a few error messages are displayed repeatedly. They scroll so fast that I can hardly read them. It seems that a file can not be opened.
My self-generated kernel 2.6.39.4 can still be started. But it seems to stop after the command is displayed for mounting of my root file system.
Now I am curious on your suggestions.
Do I need to set up a “change root environment” (e. g. from within Knoppix 6.7.1) for a clarification of my unpleasant situation?
On 2011-11-17 22:36, elfring wrote:
>
> - The desktop kernel 3.1 can be started and a few error messages are
> displayed repeatedly. They scroll so fast that I can hardly read them.
You can use a live to explore the log directory and read the boot.msg file,
if it could be written.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Unfortunately, this log file is not stored after booting the mentioned kernels. The initialisation process does not get so far on my system at the moment. (I can read a few log messages from the time before activation of the current distribution release.)
Would you like to recommend any way to find out where these kernel configurations get stuck?
I imagine that another approach would be a re-installation of openSUSE 12.1 from the DVD. But I do not like the display of the repository list by the user interface from this medium. The status column shows the setting “Removed” for all entries by default. I miss also the priorities for my collection of installation sources that I fine-tuned for the desired distribution upgrade previously.
Will this GUI application be improved?
I used a combination of the tools “zypper” and “YaST software management” for the distribution upgrade on my openSUSE 11.4 system.
does that mean you had one error or more?
Yes.
A few package repositories (like “Nvidia”) did not provide a corresponding directory for openSUSE 12.1.
I updated some components for the software “zypper”, “rpm” and “YaST” before I continued in several steps with other packages. I stumbled on an error message about a missing library symbol for the zypper command. I could circumvent this dependency issue by adjusting the environment variable “LD_LIBRARY_PATH” for my administration terminal session.
I have an intrinsic dislike of the zypper dup system upgrade method, I
prefer the DVD-boot-upgrade method - which also has its problems, but not
those pesky upgrade this first, this other library is not compatible with
the running zypper…
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On 2011-11-18 13:36, elfring wrote:
>
> robin_listas;2405245 Wrote:
>> You can use a live to explore the log directory and read the boot.msg
>> file, if it could be written.
> Unfortunately, this log file is not stored after booting the mentioned
> kernels. The initialisation process does not get so far on my system at
> the moment. (I can read a few log messages from the time before
> activation of the current distribution release.)
> Would you like to recommend any way to find out where these kernel
> configurations get stuck?
Serial port logs.
Mind: a real serial port, not a usb converter. It can send the logs to
another machine.
It is a kernel debugging tool…
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
I am curious how low or big the chances are for such a try of kernel boot debugging in my local area when I have not got an appropriate computer to connect with my “damaged” Linux configuration at the moment.
On 2011-11-20 12:06, elfring wrote:
>
> robin_listas;2405828 Wrote:
>> Serial port logs
> I am curious how low or big the chances are for such a try of kernel
> boot debugging in my local area when I have not got an appropriate
> computer to connect with my “damaged” Linux configuration at the moment.
I can’t say.
The procedure works, I have used it. Whether the information it gives can
be good to find the problem, or useless, I do not know that. Another trick
is connecting a serial port printer, but nobody has one. Perhaps with a
serial port to usb converter and usb to the printer.
You say you see a few error messages fly by fast, that method is a chance
to see them.
Another method might be a video camera. If you can then see the messages
slowly…
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
I have found an approach that is easier than taking a screenshot by a camera. I’ve got the advice to try the keyboard shortcut “Ctrl+S”. This has worked for my configuration “openSUSE desktop kernel 3.1” as it stops the display repetition.
Which analysis steps would you like suggest for the following log messages?
“…
Failed to create lvm type filter
/proc/mounts: _get_sysfs_dir: fopen %s failed: No such file or directory
/proc/devices: fopen failed: No such file or directory
…”
On 2011-11-21 13:36, elfring wrote:
>
> I have found an approach that is easier than taking a screenshot by a
> camera. I’ve got the advice to try the keyboard shortcut “Ctrl+S”. This
> has worked for my configuration “openSUSE desktop kernel 3.1” as it
> stops the display repetition.
Never thought of that. It is an old trick.
> Which analysis steps would you like suggest for the following log
> messages?
>
> “…
> Failed to create lvm type filter
> /proc/mounts: _get_sysfs_dir: fopen %s failed: No such file or
> directory
> /proc/devices: fopen failed: No such file or directory
> …”
Dunno…
That’s all it says, repeated?
Weird shot in the dark: at the grub boot prompt, use F5 to change to systemV.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Yes. - Now I assume that I lost the connection to my root file system which is stored inside a LVM2 partition somehow during my Linux distribution upgrade. Am I affected by another “adventure” with logical volume management?
Do I need to repair/recover any software component that should be stored in the corresponding “initial RAM disk”?
Which is an easy way to check if the LVM kernel module should be added again to my boot configuration?
>> Dunno…
>> That’s all it says, repeated?
> Yes. - Now I assume that I lost the connection to my root file system
> which is stored inside a LVM2 partition somehow during my Linux
> distribution upgrade. Am I affected by another “adventure” with logical
> volume management?
Sorry, I’m not an LVM expert, never used it. But I suggest you try first
the trick I mentioned about systemV.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
