Upgrade issue: 11.4 to 12.1, kde failure, didn't change version

I had 11.4 running in vm and followed some online directions for upgrading from terminal.
So updated repositories and then ran zypper.
Accept all the notices and it says 1.2 gb to download, 650mb will be final used.
So everything goes fine until a package called digikam unpacks and it fails with “bad media attach point”.
I think I chose A to skip it but may have chosen I on one of them.
The next 5 fail with the same error and then it stops back at root.

So I reboot and it takes a long time 3 minutes and log in and kde fails with “crash handler can’t even handle the crash” or something, then next says "KDE can’t load becaus /tmp is out of space. So I try to figure out what there is no space and go in /tmp to delete some files.
I find the upgrade .iso thinking that would take up space. Delete that and reboot. Same error in kde. ICE will boot.

And now I check the version in etc. and it says I’m still on 11.4.
Can anyone tell me how to fix this or try again?
Did I run out of tmp space is that why it failed?
How do I make more tmp space?
The install has 28gb or so, seemed like enough.

Clear Temp Files at Boot - Blogs - openSUSE Forums

You may have to set that up at level 3
Boot to Level 3, then Yast and More… - Blogs - openSUSE Forums

And if you are in any way using 12.1 you need to boot with systemV for the /tmp to clear (there is bug in 12.1)
http://dl.dropbox.com/u/10573557/12.1_Misc/12.1_F5_sysV.jpg

So, in the middle of a complete system upgrade, i.e. moving from 11.4 to 12.1 you got some error messages, then rebooted? If so, do yourself a favor and perform a clean 12.1 install. That’s going to take less time, since we’ll never know what was upgraded and what was not. You now are in a mix of 11.4 and 12.1 packages, which (understatement) is not too good.

Yes that’s exactly what happened.
I now tried to run update from Yast and it says root is out of space.
It’s at 4.9 GB somehow.
It doesn’t look like I can change root size unless I use some complicated steps through vmware.
So I guess starting over is all I can do.
But it’s weird if I check partition sizes root is full but /tmpfs is only 15%.
Lots of room in other places. thanks

On 2012-01-27 20:06, petecha wrote:

> But it’s weird if I check partition sizes root is full but /tmpfs is
> only 15%.

tmpfs is not the /tmp directory. Boot to text mode (runlevel 3) and paste
here the output of “df -h” using code tags.


Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

On 2012-01-27 15:36, petecha wrote:
>
> I had 11.4 running in vm and followed some online directions for
> upgrading from terminal.

Exactly which instructions?

You have to boot in text mode only (level 3), make space, repeat the zypper
dup procedure till it succeeds. If something fails, paste here the exact
procedure and error messages, using code tags.

DO NOT boot to kde.


Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)

Ok I just realized that I did not have enough room to upgrade this way.
I also think when I deleted the upgrade iso the os thought it was still there.
It should have freed up space but didn’t.
I read somewhere that rm in /tmp is not a good idea.
But I only had 8gb on that vm so that should explain why this failed.
No huge deal, I deleted and will install with some more room this time.
thank you for the help.

On 2012-01-28 21:06, petecha wrote:

> I also think when I deleted the upgrade iso the os thought it was still
> there.
> It should have freed up space but didn’t.

If a file is in use when deleted, it is not really deleted.

> I read somewhere that rm in /tmp is not a good idea.

Depends how yo do it.

> But I only had 8gb on that vm so that should explain why this failed.

Could be.


Cheers / Saludos,

Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)