Your session only lasted less than 10 seconds

[seoexpert]

Hi,

I'm new to this forum. So hello everybody. I have a small problem with a Red hat distro. (I'm using ubuntu mostly). So this post is not for one of my computers. It belongs to a friend of mine. He don't have the installation disks anymore or recovery disks. The problem is that he is using this computer for an ERP software that he runs on his network (from windows machines) and he is using the machine mostly for db entries. (I don't know exactly (neither my friend does) where the program from the windows machines is saving data. - I quess MySQL or other db). Anyway he is getting the following message on the logon screen after a recent power failure in his area (I'm not sure if this caused the problems):

Your session only lasted less than 10 seconds. etc, etc, and that probably he has run out of disk space.

View details at ~/.xsession-errors file. The message on the screen shows this:

/etc/X11/gdm/PreSession/Default: Registering your session with wtmp and utmp

/etc/X11/gdm/PreSession/Default: running: /usr/bin/X11/sessreg -a -w /var/log/wtmp -U /var/run/utmp -x "/var/gdm/:0.Xservers" -h "" -1 ":0" "root"

Agent pid 3235

Not enough free disk space on /tmp

I'm able to login as root using ctrl-alt-F1 and using in the root folder the

du -sh *

I get :

5.7M bin
3.4M boot
168K dev
34M etc
227M home
8.0K initrd
51M lib
11G local
16K lost+found
16K media
8.0K misc
8.0K mnt
8.0K opt
879M proc
11M root
14M sbin
8.0K selinux
8.0K srv
0 sys
24M temp
48K tmp
2.3G usr
96G var

The total disk space is 320GB and the system only takes 160GB (while the boot files are on a second disk which is 120GB).

So I don't think that disk space is the problem here.

I googled the message and I find various solutions but non of them worked:

eg.

chmod 1777 /tmp

or

deleting .ICEauthority or chmod to 600 before deleting

or

deleting .gnome folder (it uses kde) also didn't help.

The distro info:

Linux version 2.6.9-5.EL
(bhcompile at decompose.build.redhat dot com) (gcc version 3.4.3 20041212 (Red hat 3.4.3-9.EL4))
#1 wed Jan 5 19:22:18 EST 2005

kernel \r on an \m

LSB_Version="1.3"
Red Had Enterprise Linux AS Release 4 (Nahant)

I need some help if anyone can help me to sort this out.

Thanks in advance

[yancek]

Don't know what ERP software is?
If I understand your post, your friend is running windows 7 and conects to the Red Hat computer over a network?
Does he have physical access to the Red Hat machine?
If your friend (or whoever set this up) has a LAMP server, the database files are probably saved in the /var directory.
Your df output shows the /var directory at 96GB. Often a /var directory used as a server is on a separate partition and if that is the case here then the partition may be full? Run fdisk -l command as root to get partition information and check /etc/fstab to see if there is an entry for /var. Just a guess?

[seoexpert]

Don't know what ERP software is?
If I understand your post, your friend is running windows 7 and conects to the Red Hat computer over a network?
Does he have physical access to the Red Hat machine?
If your friend (or whoever set this up) has a LAMP server, the database files are probably saved in the /var directory.
Your df output shows the /var directory at 96GB. Often a /var directory used as a server is on a separate partition and if that is the case here then the partition may be full? Run fdisk -l command as root to get partition information and check /etc/fstab to see if there is an entry for /var. Just a guess?

Hello,

Thanks for reply. You are right. He is using windows to connect to the Red hat machine. The ERP is custom made. The programmer told him that he is using postgresql db.

And yes I have the machine in my hands now and I have root access by using ctrl+alt+F1. I have run df -h and I find out that the second hard drive is full. The one that is 120G shows 0% free. So it is not the hard drive I thought it was full. On that drive is the / mounted.

So my options would be to clone the full hard disk to a bigger one? or just copy the db and install a new distro on a new machine and move the db there?

But never did any of this process before. I see that dd and rsync are suggested for cloning. Rsync is suggested as a better solution.

Can you suggest a way? Does the disk with the / have to be mounted and the new hard disk as well? Perhaps using a live cd (rescuecd) or other? In that case do I have to mount the disks? or not? (during cloning I mean).

Any suggestion is welcome


Thanks in advance.

[yancek]

I see that dd and rsync are suggested for cloning. Rsync is suggested as a better solution.

Either should work, I've never used either so can't give you any advice there. Another option would be clonezilla which is designed to do exactly what you want. If you are not familiar with it, I would suggest you read the instructions/documentation carefully before proceeding partiuclarly with regard to cloning partition to partition or disk to disk.

[seoexpert]

Either should work, I've never used either so can't give you any advice there. Another option would be clonezilla which is designed to do exactly what you want. If you are not familiar with it, I would suggest you read the instructions/documentation carefully before proceeding partiuclarly with regard to cloning partition to partition or disk to disk.

I have clonezilla and Ping. Never used them so far. I guess that will do the job (either one of them). Thanks!