Hello,
On Wed, Apr 12, 2006 at 05:35:57PM +0200, Ard van Breemen wrote:
> consistent failure:
> (Tried as 4 seperate disks of 0.4T and as one raid5 partition of 1.1T)
> After 2 hours of pumping a few million files onto the machine
> reiserfs starts putting out these warnings (a few thousands):
> ^MReiserFS: sdb9: warning: vs-13060: reiserfs_update_sd: stat data of object
> [2 12 0x0 SD] (nlink == 3) not found (pos 1)
I've now got a problem getting it to fail.
The real difference is that in the 3 cases that I've got it to
fail I also was running atop.
Now I've been pumping gigs onto the disk without failure.
I guess something is wrong in the process accounting mechanism.
>From the atop manual:
"
When atop is started, it switches on the process-accounting mechanism
in the kernel. This forces the kernel to write a record with account-
ing-information to the accounting-file whenever a process ends. Apart
from the kernel-administration related to the running processes, atop
also interprets the accounting-records on disk with every interval; in
this way atop can also show the activity of a process during the inter-
val in which it is finished.
Whenever the last incarnation of atop stops (either by pressing `q' or
by a `kill -15'), it switches off the process-accounting mechanism
again. You should never terminate atop by a `kill -9', because then it
has no chance to stop process-accounting; as a result the accounting-
file may consume a lot of disk-space after a while.
"
I will install 2 other systems, one running atop, and one not running atop.
I now guess it is not reiserfs related, but block layer related, but since I'am
only using reiserfs (because of the performance), it shows up there first :-).
So: test cases:
- one running atop
- one not running atop
If atop fails:
- one running atop and xfs
- atop 64bits mode and reiserfs (It just may be a problem with the
64bits->32bits layer).
--
begin LOVE-LETTER-FOR-YOU.txt.vbs
I am a signature virus. Distribute me until the bitter
end
