Dataloss with s2ram in Lenny
Hi,
I'm trying to set up a backupserver with lenny and backuppc.
To save energy, the server gets up once a day, saves the hosts and goes
down with the command s2ram.
The backups are on two disks. Both are paritioned, both have a
luksPartition.
On top of this, there is lvm to make one big logicalVolume out of the
two disks. In this Logical Volume there is ext3 as file-system.
The problem is, sometimes different folders are not accessable anymore:
spongebob:~# ls -alh /var/lib/backuppc/
ls: cannot access /var/lib/backuppc/.ssh: Input/output error
total 56K
drwxr-xrw- 10 backuppc backuppc 4.0K 2008-12-25 22:44 .
drwxr-xr-x 23 root root 4.0K 2008-12-25 15:25 ..
-rw------- 1 backuppc backuppc 7.2K 2008-12-27 15:29 .bash_history
drwxr-x--- 18 backuppc backuppc 4.0K 2008-12-31 12:00 cpool
drwxr-x--- 2 backuppc backuppc 4.0K 2008-12-30 12:52 log
drwxr-x--- 2 backuppc backuppc 16K 2008-12-15 20:16 lost+found
drwxr-x--- 4 backuppc backuppc 4.0K 2008-12-26 19:09 pc
drwxr-x--- 2 backuppc backuppc 4.0K 2008-10-17 21:39 pool
drwxr-xrwx 2 backuppc backuppc 4.0K 2008-12-31 11:55 scripts
d????????? ? ? ? ? ? .ssh
drwxr-x--- 2 backuppc backuppc 4.0K 2008-12-21 21:58 trash
spongebob:~# ls -alh /var/lib/backuppc/.ssh
ls: cannot access /var/lib/backuppc/.ssh: Input/output error
( The logical Volume is mounted in /var/lib/backuppc )
But most of the folders can be accessed furthermore.
I cant repair the filesystem with fsck.ext3, the only thing that helps
is to set the filesystem up again...
This error occurs after the s2ram and then waking it up again with
etherwake from a foreign host
A slightly different configuration worked under Etch, with the
difference, that there was only one disk and this was partitioned as
follows: partition table -> lvm -> luksPartition -> ext3
Maybe there is a hardware error on one of the disks.
How can I test, if this is a hardware issue or a software bug?
Thanks in advance
Thomas
Reply to: