Plasma error when I VNC to my computer

OK I have no physical access to this computer so if I need a GUI I use VNC to access it. I use this as a web server and host teamspeak.

Basically I join mydomain.com:2 to access the box and just put in my password. Tonight when i logged in there was a popup box with something about a plasma error that closed too quick to read the rest. I am left with a black screen with an X now. I did a reboot on the server and its the same black screen, no error this time. What can I do to get my desktop back ?

I realise this is a very vague question, I am a noob when it comes to VNC but if you can point me in the right direction, kick me if you have to id appreciate it.

Well, things don’t change by their own. Any recent updates/upgrades?
Can you access the machine through ssh?

I havent done updates for about 2 months. I have access VNC several times since I did the update. Yes I have SSH/FTP access to it.

Remove / rename the plasma* files in ~/.kde4/share/config on the remote. Then login, the desktop should have the default openSUSE layout.

I only had for plasma* files, I renamed them with a .bak and its still the same black screen when I logon to domain.com:2

plasma-appletsrc.bak
plasma-desktop-appletsrc.bak
plasma-desktoprc.bak
plasmarc.bak

Any more ideas ?

When I just logged in now I got this message too
http://clanwta.com/handy/kdess.jpg

Thanks for your help bro, what you told me made that error pop up else I would never have fixed it. It works now.
I basically cleared /var/log and the /tmp/ directory. Is there a cron you can recommend to do this weekly, one that wont screw everything up ?

Glad you’ve got it solved/fixed. Normally it’s dangerous to clean out these folders by hand on a running system. The proper way to arrange for cleaning out temporary folders is:
Yast - System - Sysconfig editor, search for “TMP”, and set “CLEAR_TMP_DIRS_AT_BOOTUP” to “yes”. Confirm, reboot and /var/tmp and /tmp will be cleaned right before writing new data to them.

But this doesn’t work anymore with systemd. Those variables are just ignored now.

Please see: openSUSE 12.3 Release Notes
The default is to clean /tmp every 10 days and /var/tmp/ every 30 days.

You’re completely right. Which leaves just one other option: flooding of /var/log. And that should have a cause…

The same message just popped up and uppon a reboot the service seems like its not starting, I cant connect at all. Just said /tmp/ is full and kde wont start if I recall correctly before the screen went blank.

Space wise this is what I have below, sda2 is almost maxed, sda3 is pretty much empty. Any ideas on what to do next ?

Filesystem Size Used Avail Use% Mounted on
rootfs 20G 18G 958M 95% /
devtmpfs 988M 188K 987M 1% /dev
tmpfs 993M 27M 966M 3% /dev/shm
/dev/sda2 20G 18G 958M 95% /
/dev/sda3 437G 892M 414G 1% /home

Well, 958M free space should be more than enough to start KDE.
But try to delete everything in /tmp, /var/tmp and /var/log (or move the logfiles from /var/log to sda3 if you want to keep them)

On 09/25/2013 01:06 PM, wolfi323 wrote:
>
> originalhandy;2587608 Wrote:
>> The same message just popped up and uppon a reboot the service seems
>> like its not starting, I cant connect at all. Just said /tmp/ is full
>> and kde wont start if I recall correctly before the screen went blank.
>>
>> Space wise this is what I have below, sda2 is almost maxed, sda3 is
>> pretty much empty. Any ideas on what to do next ?
>>
> Well, 958M free space should be more than enough to start KDE.
> But try to delete everything in /tmp, /var/tmp and /var/log (or move the
> logfiles from /var/log to sda3 if you want to keep them)

I have not seen what filesystem you have on /dev/sda2, but if it is ext3 or
ext4, the standard setup allows 5% of the fs to be reserved only for root. On
your 18 GB partition, that amounts to 0.9 GB, which sounds suspiciously like
your remaining free space. You can verify the amount with “sudo tune2fs -l
/dev/sda2” and look at the “Reserved block count” line. The -m option to tune2fs
lets you change the percentage, and the -r option allows you to specify a new
block count. See “man tune2fs”.

I empited those tmp and its still the same.

I cleared the tmp logs etc and I am “up” to this

Filesystem Size Used Avail Use% Mounted on
rootfs 20G 18G 1.2G 95% /
devtmpfs 988M 188K 987M 1% /dev
tmpfs 993M 27M 966M 3% /dev/shm
/dev/sda2 20G 18G 1.2G 95% /
/dev/sda3 437G 892M 414G 1% /home

Is 18GB normal for something like this ? Its a web server with no website on it right now and teamspeak. Nothing else is done with it.

sudo tune2fs -l /dev/sda2 came back with “tune2fs: command not found”

Well, this depends on how much software you have installed. But 18GB on / sounds a bit much for just a web server.
Maybe you have a lot of packages in /var/cache/zypp/packages? This can be deleted as well without doing any harm.
Try to run “du -h /” to find out where most of your space is wasted.

sudo tune2fs -l /dev/sda2 came back with “tune2fs: command not found”

Try “sudo /usr/sbin/tune2fs -l /dev/sda2”.
If that doesn’t work either you may have to install the package “e2fsprogs”.

Here’s the output from my laptop:


# df -h
Bestandssysteem Grootte Gebruikt Besch Geb% Aangekoppeld op
devtmpfs           2,9G     8,0K  2,9G   1% /dev
tmpfs              2,9G     928K  2,9G   1% /dev/shm
tmpfs              2,9G     4,0M  2,9G   1% /run
**/dev/sda1           21G      19G  866M  96% /**
tmpfs              2,9G        0  2,9G   0% /sys/fs/cgroup
tmpfs              2,9G     4,0M  2,9G   1% /var/run
tmpfs              2,9G     4,0M  2,9G   1% /var/lock
/dev/sda5          4,8G     2,5G  2,1G  56% /srv
/dev/sda6           50G      44G  3,1G  94% /home
/dev/sda7           36G      25G  8,7G  74% /Extra



Note: this is my sandboxing laptop. It has KDE, GNOME, LXDE, Xfce, webserver (/srv is mounted on it’s own partition), both Libreoffice and Calligra, a couple of kernels ( I keep three atm ). But if yours is a KDE only, a GNOME only and no huge games have been installed ( you can check whether “large” packages have been installed by playing with the sorting order of packages in Yast ), no huge databases have been stored. then something else is going on. This shows the folders on the root with their size:


cd /
su -c 'du -h --max-depth=1'

It is literally a stock install except for TS3, XAMPP and thats about it. TS3 is about 5MB I think, not big at all. I have a few meg bot that runs it too.

administrator@ws-19476:/> su -c 'du -h --max-depth=1'Password:
1.4M    ./srv
4.0K    ./mnt
29M     ./etc
4.0K    ./selinux
24M     ./boot
730M    ./opt
13M     ./sbin
8.0K    ./.config
du: cannot access `./var/lib/ntp/proc/5461/task/5461/fd/4': No such file or directory
du: cannot access `./var/lib/ntp/proc/5461/task/5461/fdinfo/4': No such file or directory
du: cannot access `./var/lib/ntp/proc/5461/fd/4': No such file or directory
du: cannot access `./var/lib/ntp/proc/5461/fdinfo/4': No such file or directory
14G     ./var
2.2M    ./root
605M    ./home
88K     ./tmp
8.3M    ./bin
du: cannot access `./proc/5461/task/5461/fd/4': No such file or directory
du: cannot access `./proc/5461/task/5461/fdinfo/4': No such file or directory
du: cannot access `./proc/5461/fd/4': No such file or directory
du: cannot access `./proc/5461/fdinfo/4': No such file or directory
0       ./proc
3.1G    ./usr
29M     ./dev
0       ./sys
148M    ./lib
16K     ./lost+found
4.0K    ./media
18G     .
administrator@ws-19476:/>



I will run “du -h /” and “sudo /usr/sbin/tune2fs -l /dev/sda2” as soon as I get home. I only had time to log on real quick before I had to split for work again.

Do I just delete packages in /var/cache/zypp/packages? or is there a certain command I need to do to do it safely ?

No need to.
So most of your missing space is in /var (14G).
Try to run “du -h /var” (maybe with “–max-depth=1”) then to get further information where it is.

Do I just delete packages in /var/cache/zypp/packages? or is there a certain command I need to do to do it safely ?

Just remove the directory. That is safe.

I removed the directory and this is where I am at

ws-19476:/ # df -h
Filesystem            Size  Used Avail Use% Mounted on
rootfs                 20G   16G  3.1G  84% /
devtmpfs              988M  188K  987M   1% /dev
tmpfs                 993M   46M  948M   5% /dev/shm
/dev/sda2              20G   16G  3.1G  84% /
/dev/sda3             437G  803M  414G   1% /home
ws-19476:/ #


Is there anything else I need to do or look at ? Is there a way to change partician sizes with ssh or does it have to be with external software ? If I need to use external, whats the safest to use ? I can get a guy on site to do it.

I uploaded my (compressed site) and am left with

administrator@ws-19476:~> df -h
Filesystem            Size  Used Avail Use% Mounted on
rootfs                 20G   17G  2.6G  87% /
devtmpfs              988M  188K  987M   1% /dev
tmpfs                 993M   45M  949M   5% /dev/shm
/dev/sda2              20G   17G  2.6G  87% /
/dev/sda3             437G  803M  414G   1% /home
administrator@ws-19476:~>


Now on restart the server wont run at all when I connect to domain.com:2 I tried vncserver restart and nada.