Hi,
I know I’m supposed to upgrade gradually from 13.1 > 13.2 > 42.1 but was just wondering if I just get the ISO and do an offline upgrade from my 13.1 to 42.1, is it so risky? Anyone did this?
I don’t have any fancy LVM or advanced configs.
less risky way. But it is a large leap:P
Backup anything you want to be sure to keep.
You may or may not have problem with some config files in you home
If you back up your HD first (use Clonezilla, for example), the risk will be mitigated. You could then try it (13.1 > 42.1), and if it does not work out well, just restore from the backup and try again with the 13.1 > 13.2 > 42.1 full steps.
Well, I did upgrade it and offcourse it failed, I got the “error: premature end of file /vmlinuz” in Grub, failing to boot.
So I went and booted up the rescue disk, mounted the system and reinstalled grub2 using this tutorial here:
https://forums.opensuse.org/content.php/128-Re-install-Grub2-from-DVD-Rescue
Be aware, the commands for mounting the /proc and /sysfs are
"mount -t proc proc /proc" instead of "mount /proc"
and
"mount -t sysfs sysfs /sys" instead of "mount /sys"
Anyways, I got myself running on 13.2 and decided to download the 42.1 ISO and make a bootable USB drive so I can do the upgrade offline, but I get stuck at the partition steps, I get an error after I select my 13.2 installation partition and click Next, see the picture.
Now what?
Image here http://imgur.com/0BGYkCa
My drives are like this
sda - HDD 160GB with one partition sda1 used for misc data
sdb - SSD 120GB with sdb1 as swap, sdb2 as root, sdb3 as my /home/me
Here is a dump of fdisk-l
Disk /dev/sda: 149.1 GiB, 160040803840 bytes, 312579695 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0xed46604b
Device Boot Start End Sectors Size Id Type
/dev/sda1 2048 312578047 312576000 149G 83 Linux
Disk /dev/sdb: 111.8 GiB, 120034123776 bytes, 234441648 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0xd420d420
Device Boot Start End Sectors Size Id Type
/dev/sdb1 2048 4208639 4206592 2G 82 Linux swap / Solaris
/dev/sdb2 * 4208640 46153727 41945088 20G 83 Linux
/dev/sdb3 46153728 210270207 164116480 78.3G 83 Linux
Here is my fstab
UUID=76ec4d66-ed02-4b07-bb1b-51f90dc49446 swap swap defaults
0 0
UUID=a7f5917d-dba0-4d89-9a6f-79c4547ea314 / ext4
acl,user_xattr,noatime 1 1
UUID=58f62fa4-e31d-4b59-9b44-545b91b769cf /home ext4 acl,user_xattr,noatime
1 2
UUID=9f583edd-d181-48f6-959d-087b931194ba /home/roberto/Data ext4 acl,user_xattr,noatime
I can see the root on the second line though. Is this related to my fstab using “by UUID”?
If you changed the partitions they may not be labelled the same. Ie if you resize/move/change then the partition is in essence recreated and gets a different UUID
So change to expert mode and point things to the correct partitions. Apparently somewhere along the way you changed something and the fstab does not reflect it
Note in expert mode be sure that the installer does not guess you want to format root. To upgrade you do not want to format root.
I managed to fix it.
I opened up /etc/fstab with nano and just re-wrote the lines, for some reason they had extra spaces, basically all over the place, and I just made them nicer and also I noticed that my last entry for the extra misc 160GB HDD did not have 0 0 parameters at the end, so I added those, saved the file, did a reboot, and voila, 42.1 saw the partition right on.
Here is my etc/fstab now
UUID=76ec4d66-ed02-4b07-bb1b-51f90dc49446 swap swap defaults 0 0
UUID=a7f5917d-dba0-4d89-9a6f-79c4547ea314 / ext4 acl,user_xattr,noatime 1 1
UUID=58f62fa4-e31d-4b59-9b44-545b91b769cf /home ext4 acl,user_xattr,noatime 1 2
UUID=9f583edd-d181-48f6-959d-087b931194ba /home/roberto/Data ext4 acl,user_xattr,noatime 0 0
Writing from openSUSE 42.1 Leap 
After upgrading I had some issues with it, there was no GUI and several error windows related to Plasma were showing up. I tried the recovery option in Grub but still same thing. I knew it has to be the nVidia drivers, so I switched to a console with ctrl+alt+f1, made myself root and added the nVidia repo for 42.1 Lead, then I did a zypper up in order to re-install the nvidia packages. It took a while and some weird thing went on, basically the screen was scrolling with thousands of unknown characters for like…10 minutes. Eventually it stopped saying something about some kernel modules. After rebooting I now have my working 42.1 KDE working.
When you upgrade some old stuff gets left behind and can cause odd thing but seem you did get a working machine. I recommend you get all updates there have been a bundle of fixes
Well done - your bravery may be of help to others who want to take the risk, but as this account shows it’s not a trivial exercise.
I’m now experiencing problems with opening up Chromium and other apps, I keep getting this error about Kdeinit could not lauch ksystraycmd
See here
I tried reinstalling Chromium several times, it works fine if I run it via terminal but not via click or by clicking links in any other app.
Couldn’t find very much on the web about this 
No, Plasma 5 does not support the former KDE4 system tray icons IIRC, so you’ll need to recreate the links to your apps I think. I wouldn’t be surprised if you encounter other problems related to stale/legacy configuration files because of your upgrade method. All part of the learning though.
BTW, are you able to launch apps via the menu? If so, just right-click on a given app icon and ‘Add to Panel’
So far, only Chromium is giving me headaches, I can run any app via the menu, except Chromium. Add to panel, adds the icon but on click, it gives the same error.
I also went in Yast and configured default internet app to be Chromium, but still no luck.
Try editing the menu, right click ‘Edit Applications’… does that lead to a working icon?
Hi,
Icon is and was ok, I tried that, it justs open the edit menu, everything seems ok.
I really don’t know what should I do anymore. I need Chromium, and I have to run it via terminal every time.
I’ve never used chromium, but in an effort to assist I just installed using
# zypper in chromium
Loading repository data...
Reading installed packages...
Resolving package dependencies...
The following 3 NEW packages are going to be installed:
chromium chromium-desktop-kde chromium-ffmpeg
The following recommended package was automatically selected:
chromium-ffmpeg
Does this reflect the packages you have installed?
Once installed, I then went to the menu and launched it with no issues. Then menu entry contains ‘chromium %u’
Yes, I had it exactly like that.
I also ran a “zypper remove chromium” the reinstalled it again with “zypper in chromium”, exactly the same problem.
Yes, menu entry contains ‘chromium %u’.
This is the output from starting chromium using the terminal, the only way I can launch it.
roberto@linux-jspa:~> chromium
[11540:11566:0106/195527:ERROR:logging.h(810)] Failed to call method: org.kde.KWallet.isEnabled: object_path= /modules/kwalletd: org.freedesktop.DBu
s.Error.NoReply: Message did not receive a reply (timeout by message bus)
[11540:11566:0106/195527:ERROR:native_backend_kwallet_x.cc(396)] Error contacting kwalletd (isEnabled)
[11540:11566:0106/195527:ERROR:logging.h(810)] Failed to call method: org.kde.KWallet.isEnabled: object_path= /modules/kwalletd: org.freedesktop.DBu
s.Error.NoReply: Message did not receive a reply (timeout by message bus)
[11540:11566:0106/195527:ERROR:native_backend_kwallet_x.cc(396)] Error contacting kwalletd (isEnabled)
[11540:11566:0106/195537:ERROR:generic_change_processor.cc(383)] Autofill datatype error was encountered: Failed to delete Autofill node. Local data
. ActOnChanges@../../components/autofill/core/browser/webdata/autocomplete_syncable_service.cc:396could not find entry matching the lookup criteria.
[11540:11566:0106/195537:ERROR:generic_change_processor.cc(322)] Delete: Bad entry.
[11540:11566:0106/195537:ERROR:shared_change_processor.cc(109)] Change processor disconnected.
libpng warning: iCCP: known incorrect sRGB profile
libpng warning: iCCP: known incorrect sRGB profile
I tried again removing Chromium and also manually remove the chromium folder under /home/me/.config then reinstalled Chromium fresh, no luck. Still, the same stupid error.
:(>:(
Well, Plasma 5 doesn’t include ksystraycmd AFAIU. It requires the legacy xembed tray protocol which is not supported by Plasma 5, so the error you reported in post #11 doesn’t make sense to me. I’m not seeing this erorr when launching from the KF5 menu. Hopefully others can comment further here.
Can you post your configured repos?
zypper lr -d
Okay, I think I’ve figured out what might occurring here. I can replicate the error you report if the menu item for Chromium has ‘Place in system tray’ enabled. Uncheck that option and save when done. You should then find Chromium launches from the menu as expected.