Hello altogether,
since a week I am facing a serious issue with my filesystems. After boot and log-in into Gnome or KDE, the harddisk starts working heavily even though programmes and data are loaded to RAM. My only harddisk access then is the normal web-cache and work with some documents. After this activity my two filesystems root and home often become flagged as read-only. I then cannot modify my files of course and cannot access root privileges
Starting a programme from shell I get this error: sudo: unable to open /var/lib/sudo/simon/0:root: filesystem is read-only but it’s executed (with a lot of obvious errors).
Stating a programme with root privileges from the desktop such as yast2 I get the following log-message: kernel: 1602.299884] journal commit I/O error.
I had this problem the first time a week ago, restarted the computer and everything was fine again. Today even five reboots didn’t help for more than a minute or two. I’m now using opensuse since version 11.2 almost daily on various computers and never had such a problem before :(.
These are my system-specs:
opensuse 12.1 x86_64
kernel desktop 3.5.1 (after recommendation due to some serious memory leaks regarding my trackpad and graphics-driver)
AMD catalyst 12.6 (12.4?) with newest flgx (which finally fixed some annoying bugs)
ASUS Eee-PC 1215
AMD E-450 Brazos
AMD Radeon 6320 (included)
Any suggestion for a fix? Or better trying directly the hard way with a complete fresh install?
Many thanks in advance!
Simon
On 2012-08-06 16:36, SimonHeimbach wrote:
>
> Hello altogether,
>
> since a week I am facing a serious issue with my filesystems. After
> boot and log-in into Gnome or KDE, the harddisk starts working heavily
> even though programmes and data are loaded to RAM. My only harddisk
> access then is the normal web-cache and work with some documents. After
> this activity my two filesystems root and home often become flagged as
> read-only. I then cannot modify my files of course and cannot access
> root privileges
…
> Any suggestion for a fix?
Check the logs.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
Well, here we are, a short excerpt of the logs I find a bit suspicious:
Aug 6 18:36:29 linux-9t6q simon: Try 5/8 : /proc/fs/vmblock/dev not available. sleeping for 32 seconds
Aug 6 18:37:01 linux-9t6q simon: Try 6/8 : /proc/fs/vmblock/dev not available. sleeping for 64 seconds
Aug 6 18:38:05 linux-9t6q simon: Try 7/8 : /proc/fs/vmblock/dev not available. sleeping for 128 seconds
Aug 6 18:42:23 linux-9t6q kernel: 453.732419] ata1.00: exception Emask 0x0 SAct 0x3 SErr 0x0 action 0x0
Aug 6 18:42:23 linux-9t6q kernel: 453.732433] ata1.00: irq_stat 0x40000008
Aug 6 18:42:23 linux-9t6q kernel: 453.732443] ata1.00: failed command: READ FPDMA QUEUED
Aug 6 18:42:23 linux-9t6q kernel: 453.732463] ata1.00: cmd 60/00:00:38:12:70/01:00:00:00:00/40 tag 0 ncq 131072 in
Aug 6 18:42:23 linux-9t6q kernel: 453.732463] res 41/40:00:cf:12:70/00:01:00:00:00/00 Emask 0x409 (media error) <F>
Aug 6 18:42:23 linux-9t6q kernel: 453.732471] ata1.00: status: { DRDY ERR }
Aug 6 18:42:23 linux-9t6q kernel: 453.732477] ata1.00: error: { UNC }
Aug 6 18:42:23 linux-9t6q kernel: 453.769396] ata1.00: configured for UDMA/133
Aug 6 18:42:23 linux-9t6q kernel: 453.769443] sd 0:0:0:0: [sda] Unhandled sense code
Aug 6 18:42:23 linux-9t6q kernel: 453.769451] sd 0:0:0:0: [sda]
Aug 6 18:42:23 linux-9t6q kernel: 453.769456] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Aug 6 18:42:23 linux-9t6q kernel: 453.769463] sd 0:0:0:0: [sda]
Aug 6 18:42:23 linux-9t6q kernel: 453.769468] Sense Key : Medium Error [current] [descriptor]
Aug 6 18:42:23 linux-9t6q kernel: 453.769478] Descriptor sense data with sense descriptors (in hex):
Aug 6 18:42:23 linux-9t6q kernel: 453.769483] 72 03 11 04 00 00 00 0c 00 0a 80 00 00 00 00 00
Aug 6 18:42:23 linux-9t6q kernel: 453.769508] 00 70 12 cf
Aug 6 18:42:23 linux-9t6q kernel: 453.769520] sd 0:0:0:0: [sda]
Aug 6 18:42:23 linux-9t6q kernel: 453.769527] Add. Sense: Unrecovered read error - auto reallocate failed
Aug 6 18:42:23 linux-9t6q kernel: 453.769534] sd 0:0:0:0: [sda] CDB:
Aug 6 18:42:23 linux-9t6q kernel: 453.769538] Read(10): 28 00 00 70 12 38 00 01 00 00
Aug 6 18:42:23 linux-9t6q kernel: 453.769558] end_request: I/O error, dev sda, sector 7344847
Aug 6 18:42:23 linux-9t6q kernel: 453.769594] ata1: EH complete
Hope that helps.
Filesystem type is ext4.
Thank you.
On 2012-08-06 19:16, SimonHeimbach wrote:
>
> Well, here we are, a short excerpt of the logs I find a bit suspicious:
You should have used code tags.
>> Aug 6 18:36:29 linux-9t6q simon: Try 5/8 : /proc/fs/vmblock/dev not available. sleeping for 32 seconds
>> Aug 6 18:37:01 linux-9t6q simon: Try 6/8 : /proc/fs/vmblock/dev not available. sleeping for 64 seconds
>> Aug 6 18:38:05 linux-9t6q simon: Try 7/8 : /proc/fs/vmblock/dev not available. sleeping for 128 seconds
>> Aug 6 18:42:23 linux-9t6q kernel: 453.732419] ata1.00: exception Emask 0x0 SAct 0x3 SErr 0x0 action 0x0
>> Aug 6 18:42:23 linux-9t6q kernel: 453.732433] ata1.00: irq_stat 0x40000008
>> Aug 6 18:42:23 linux-9t6q kernel: 453.732443] ata1.00: failed command: READ FPDMA QUEUED
You do have problems with the hard disk, but I can not say what. It can be the cause of the
read-only problem, there could be some mention of that in the logs - if the note can be written. If
not, you might see it in VT10.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 “Asparagus” GM (bombadillo))
Hi!
last night I did a complete reinstall of the system. As I bought the computer with opensuse pre-installed, I also had a few little issues that I directly got rid of as well.
Everything now works fine.
My guess is that the journal of the filesystem got corrupted when the battery went low during stand-by. Correct me if I’m wrong. And let’s hope it won’t occur again.
Anyway, thank you for your advice.
Simon
On 2012-08-07 23:26, SimonHeimbach wrote:
>
> Hi!
>
> last night I did a complete reinstall of the system. As I bought the
> computer with opensuse pre-installed, I also had a few little issues
> that I directly got rid of as well.
>
> Everything now works fine.
>
> My guess is that the journal of the filesystem got corrupted when the
> battery went low during stand-by. Correct me if I’m wrong. And let’s
> hope it won’t occur again.
Doubtful, but… Perhaps, if there are bad sectors they get remapped on write by the installation
procedure.
I think you should run the SMART long test. It is done with smartctl
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 “Asparagus” GM (bombadillo))