You are not logged in.
My 64bit system was suddenly rendered unusable and I eventually got a message about the /tmp directory being out of space. I thought the problem would be cleared by a reboot, but what happened was that KDE could not be loaded due to lack of space...
My separate /home partition should have at least 20 GiB free space, and I have 2 GB RAM so, as a Linux illiterate, I don't understand what happened. I have tried to look into /tmp from the command line and found three directories, but no visible/deletable files.
If the problem should be that my root partition has somehow been saturated, which files would you suggest to look for to erase? Or do you have a different approach to this this problem? I am lost and without my favourite system...
Also, the system complains that it can no longer find and load "rlocate". It used to be there, but has apparently disappeared during an upgrade.
Any help greatly appreciated!
Last edited by whaler (2008-08-24 08:16:27)
Offline
Look in /var/log and see if a log file got out of control. If it has, take a note of the repeated error message before clearing it so you can investigate it later.
Offline
Look in /var/log and see if a log file got out of control. If it has, take a note of the repeated error message before clearing it so you can investigate it later.
Thanks for the tip. Will investigate...
Offline
Your tip seems to have solved the problem - I found that everything.log, daemon.log and errors.log all had the size of 1.3 GiB!!
Trying to look at one of them with nano resulted in an unresponsive system - seems nano may not be capable of loading monster files... I decided to just erase them. I expect Arch will generate new ones when necessary.
Things seem to be back to normal, except that dmesg reports that a volume I use to mount is now unrecognizable. The temporary lack of disk space may have caused this, or there may be an underlying hardware problem causing both... Is there a known common cause for lost codepages?
Anyway - thanks a lot for the help!
Offline
Your tip seems to have solved the problem - I found that everything.log, daemon.log and errors.log all had the size of 1.3 GiB!!
Trying to look at one of them with nano resulted in an unresponsive system - seems nano may not be capable of loading monster files... I decided to just erase them. I expect Arch will generate new ones when necessary.
Things seem to be back to normal, except that dmesg reports that a volume I use to mount is now unrecognizable. The temporary lack of disk space may have caused this, or there may be an underlying hardware problem causing both... Is there a known common cause for lost codepages?
Anyway - thanks a lot for the help!
Check out logrotate. It will stop this problem and can be tuned for your machine. It's in Core.
Last edited by skottish (2008-08-24 05:09:53)
Offline
Thanks, skottish, will do!
Meanwhile I found that I could mount all partitions except one - one which was mounted when the accident happened. Since it is the home partition of another OS, all that was needed to fix it was to boot that OS and let chkdsk do it's automatic repair job. I mention this just in case others experience something similar.
Last edited by whaler (2008-08-24 15:55:56)
Offline