System:I/O Errors

[genoot]

We run a simple DSS V6 with 16x 1TB disks for backups, connected to an Areca 1222 Raid controller. (Raid 5)
It is accessed with nfs and smb from different clients.
Since a few days, the system freezes regularly when data is copied to the storage.
Only a reset of the computer helps (for a short time)

errors are like (same found in critical_erros.log)
kernel:[11951.615468] scsi cmnd aborted, scsi_cmnd(0xffff88007fa2fe00), cmnd[0x8a,0x 0,0x 0,0x 0,0x 0,0x 2,0xae,0xdd,0x83,0xc0,0x 0,0x 0], scsi_id = 0x 0, scsi_lun = 0x 0.

and the last hour also this urgen message :
System:I/O Errors detected on unit S001. The unit requires your urgent attention in order to decrease the risk of data loss


I checked :
- we have the latest Firmware for our Areca RAID controller
- DSS software is updated to latest version b6156
- Logical volume is checked through RAID controller : no errors found.

Checking the disk from Console Tools isn't really helpfull : it complains about filesystems being mounted...

Any ideas what I should do next ?

[michi616]

use you the light version of DSS V6?
Have you a iscsi lun configurated?

[genoot]

No, we run a full version.
And no iscsi lun configurated.
Just some NAS shares.

use you the light version of DSS V6?
Have you a iscsi lun configurated?

[genoot]

it doesn't look any better 1 day later.
Not 1 share is available...


Warning
The path to this share does not exist. It is impossible to access this share. You can create the folder the share was pointing to e.g. by using the browser below. If the original folder is a snapshot, it is now inactive. If the original folder is a volume which serves as destination for volume replication, it cannot be accessed directly (only via a snapshot, please consult the online help for more details).

No, we run a full version.
And no iscsi lun configurated.
Just some NAS shares.

[Gr-R]

Have you run the file system repair from the console?
http://kb.open-e.com/File-system-repair_138.html

[genoot]

Yes, I did without any result.

The vendor told me the PCI bus on the motherboard might be causing the problem.
Motherboard and RAID controller will be replaced.

Have you run the file system repair from the console?
http://kb.open-e.com/File-system-repair_138.html

[Maciej Dytkowski]

Genoot,

Check your RAID consistency - from my experience timed out scsi commands can point to faulty drive or backplane.