LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - General (http://www.linuxquestions.org/questions/linux-general-1/)
-   -   File System Corruption - Causes and Consequences? (http://www.linuxquestions.org/questions/linux-general-1/file-system-corruption-causes-and-consequences-490404/)

Ingla 10-07-2006 02:44 PM

File System Corruption - Causes and Consequences?
 
Hi.

I'm running Breezy (Ubuntu 5.10). Lately, I've been expriencing unexplained total freezes which leave me no option but to hit the button (no mouse or keyboard). These are not associated with any particular action on my part, but may be caused by one or more programs ... or having them running together. There is also no regular time period. Breezy may run fine all night or may freeze after just a few minutes and keep doing it.

I haven't figured out what's causing this, but I'm concerned about some of the apparent results.

Specifically, the file system gets messed up fairly regularly. On boot, I frequently (sometimes after 30 boots and sometimes because of corruption in the meantime) get kicked into a console running fsck. The results are usually similar. Fsck fails and needs to be run manually.

The readout looks something like this (not taken from my machine, but very similar):

--------------------------------------------------------------------------
fsck 1.38 (30-Jun-2005)
e2fsck 1.38 (30-Jun-2005)
Backing up journal inode block information.

PALFA_A103 contains a file system with errors, check forced.
Pass 1: Checking inodes, blocks, and sizes

Running additional passes to resolve blocks claimed by more than one inode...
Pass 1B: Rescanning for multiply-claimed blocks
Multiply-claimed block(s) in inode 4133: 78675968 78675970 78675972 78675973 78675974 78675977 78675978 78675979 78675982 78675983 78675985 78675986 78675987 78675989
Pass 1C: Scanning directories for inodes with multiply-claimed blocks.
Pass 1D: Reconciling multiply-claimed blocks
(There are 1 inodes containing multiply-claimed blocks.)

File /p2030/p2030.G48.75-00.17.C.wapp3.53910.0091 (inode #4133, mod time Sat Jun 24 02:16:36 2006)
has 14 multiply-claimed block(s), shared with 1 file(s):
<filesystem metadata>
Free blocks count wrong for group #0 (24533, counted=24519).
Fix<y>? yes

Free blocks count wrong for group #2401 (14795, counted=14809).
Fix<y>? yes


PALFA_A103: ***** FILE SYSTEM WAS MODIFIED *****
PALFA_A103: 280/95392 files (8.9% non-contiguous), 92570714/97677200 blocks
-------------------------------------------------------------------------

Actually, my system contains a larger number of mulitply claimed blocks in several inodes. And it requires more cloning and fixing.

I answer "yes" to all requests to clone and fix, and then reboot. That makes everything fine until the next time.

HOWEVER, each time I see that the percentage of non-contiguous files grows by about .1%. At present, it's only 2.2%, but I'm wondering where this will lead and whether anything can be done to fix it.

Under Windows, I have programs which will make everything contiguous, but have no idea whether or how I can do that on Linux. Sooner or later, I assume this will begin to affect performance more and more.

Can anyone help me understand what's happening and especially what, if anything, I can do about it?

Any guidance much appreciated.

unSpawn 10-10-2006 08:09 AM

Ext2/ext3 defragmentation has never been a an issue. Basically it takes care of itself.


Lately, I've been expriencing unexplained total freezes which leave me no option but to hit the button (no mouse or keyboard). These are not associated with any particular action on my part, but may be caused by one or more programs ... or having them running together.
Could be anything from bad RAM to overheating to memory leaks to plain running out of memory + swap. Anything. I suggest you start by looking for clues in your syslog (see /etc/syslog.conf for details) and listing your hardware specs.

Ingla 10-10-2006 10:53 AM

Hi.

Thanks for the reply.

I ran memtest86. Got all kinds of bad addresses and errors about 40% through the first pass. So, it looks like your very first idea might be it...memory trouble. I'll change the chip before looking further.

Thanks.

Gethyn 10-10-2006 11:59 AM

I had a similar problem a few weeks back. Turns out the PSU was overloaded, I upgraded it and the problem stopped happening.


All times are GMT -5. The time now is 11:11 AM.