10.04 Login Loop. VM. Supern00b, please advise.

[theunfrailhale]

Hey gents. I have a VM of 10.04 that is running a library management software set called KOHA.

On occasion, when I reboot this VM, instead of the usual medium brown background at login, it is black instead, and It puts me into a login-loop, where I enter my credentials, and it blips and returns me to the login page.

I have been referring to this common problem in this forum post here:
http://ubuntuforums.org/showthread.p...473152&page=2&

I was in the process of attempting the solution as follows

This is how I fixed this problem on my computer:


During the boot up, hit ctrl-alt-F1 for a tty login

Use these commands:

sudo stop gdm
sudo apt-get purge gdm
sudo apt-get install wdm
sudo reboot now

all things were going swell till I ran into the error

Package wdm is not available, but is referred to by another package. 
This may mean that the package is missing, has been obsoleted, or is only available from another source. 
E: package wdm has no installation candidate

I assume that my repositories are out of date as this flavor of Ubuntu is out of date, but I need it fixed to export the library database before I can abandon it. I ran

sudo apt-get update

and it errored on some things but not all. Still didnt fix the issues of installing wdm.

Any Ideas??? I'm pretty noob with my understanding of the inner workings of linux, but have a cursory knowledge regarding navigation, syntaxes etc.

Regarding the cause of this problem, I thought it was related to the dynamic disk size of the VM so I cloned it and re-sized it to 60GB from 40GB. that fixed a problem I was having with the machine not booting, but apparently it was only a temporary fix.

Any help would be greatly appreciated

[TheFu]

You are correct - 10.04 desktop has been unsupported since April officially. 10.04 server (without the GUI) will be supported until Aprill 2015. If you have the desktop version installed, there isn't much we can do to help. If, on the other hand, you have the server version ....

apt-get update only updates the cache of packages. It doesn't upgrade any of them. The most common method to maintain packages on a server platform is to run:
sudo apt-get update && sudo apt-get dist-upgrade
on a single line. This does both things AND will install newer kernels, which is probably what you want. Don't worry that "dist-upgrade" sounds like it will upgrade your distribution. It will not. There is a different program for that. ... however, on a server, I'd only do a 10.04 to 12.04 upgrade after making a 100% trustworthy backup. Things can go wrong and I know from personal experience that server software dependencies often break badly between releases like that.

So, run the commands above regardless of desktop/server and report back with any errors. Once the VM is as current as possible, we can help more.

BTW, I have a number of 10.04 servers running still and getting updates. We also have 12.04 servers running. Both are very stable, but many large, complex solutions take about a year to migrate between LTS to the next LTS release. During these migrations, support tool dependencies often break - for example, my Zimbra server had issues because the platform Perl provided changed and some of the perl libraries/modules changed between the distros. By waiting a few months after 12.04 was released for the Zimbra team to figure out all the new dependencies, the upgrade wasn't completely smooth, but it was possible and someone else had already figured out the new perl-whatever-whatever modules that I needed pre-installed. Much easier. I only run LTS releases and will be on 12.04 until 14.04 is out 6 months.

[theunfrailhale]

Thanks for the info TheFu, I am going to try running those commands anyways, but I'm certain it is 10.04 Desktop because it normally boots with the GUI.



I'll post what I find out regardless.
Thanks!

[theunfrailhale]

OK so the issue is the same that I was having before, the VM is running out of disk space and things are too crammed for it to boot up. ran the df command and the output for /dev/sda1 was that it was 100% used. I ran sudo apt-get clean and sudo apt-get autoremove. This bought me some more space but df reports that sda1 is still at 100%.

When I ran into this problem before, my VM hdd was full, so I migrated the image to a larger virtual HDD, but I am guessing that this didn't resize the partition for sda1, so thats what I'm gunna look into now. After I get this up and running, it would seem that I would be best off upgrading the kernel to the latest LTS.

[TheFu]

Kernel updates and distro updates are not the same thing. I think you need a distro update.

Being out of space is really, really bad on UNIX-like OSes. Everytime you ssh into the machine with over 90% disk used, a warning is displayed for a good reason. THAT is when we need to address the issue - at 90%. At 100%, we don't usually have enough working space to clean things up.

Also, with lots of small files on a small-ish partition, you might have 20% storage left, but be out of inodes. df -i will show those. Being out of inodes is worse that being out of storage. It doesn't happen much when partitions are larger than 10G, but many of my VMs here only get 4G of storage. With scripting languages, they like to use lots and lots of tiny files - each sucking up inodes. Running out happens all the time with default file system parameters.,