random Read only filesystem

bibax

New Member
May 3, 2013
28
0
1
Hi,

I have an issue with a filesystem on a centos virtual sever.
The virtual server has 2 virtual disk. One for system (50 Go) and the other for datas (19 To).
Filesystem on data is ext4 on lvm volume.

At the beginning, I had no issue because I copied 11 To on the disk without any problem.
But now, when I synchronized some data, the filesystem enter in read only mode. Only vdb (data disk) is like that.
I can write on system disk but not on data disk...
I have done some e2fsck and fsck repair and some reboot but always the issue when I have to write some data on disk volume.
This is the log in message when it arrives :
Mar 4 09:54:48 sauvegarde kernel: end_request: I/O error, dev vdb, sector 338069248
Mar 4 09:54:48 sauvegarde kernel: EXT4-fs warning (device dm-2): ext4_end_bio:332: I/O error -5 writing to inode 154337303 (offset 243269632 size 8388608 starting block 42258400)
Mar 4 09:54:48 sauvegarde kernel: Buffer I/O error on device dm-2, logical block 42258400
Mar 4 09:54:48 sauvegarde kernel: Buffer I/O error on device dm-2, logical block 42258401
Mar 4 09:54:48 sauvegarde kernel: Buffer I/O error on device dm-2, logical block 42258402
Mar 4 09:54:48 sauvegarde kernel: Buffer I/O error on device dm-2, logical block 42258403
Mar 4 09:54:48 sauvegarde kernel: Buffer I/O error on device dm-2, logical block 42258404

The server is a centos 7 up to date and the promox is in version 3-4.11.
The raid 6 for proxmox server is ok and all disk seems to be fine.
I don't know how to solve this issue...
Do you have any idea?

Do I have to delete the data virtual and recreate it or do you have an other idea?

Thank you in advance for your help!

Regards
 
We have an cluster with 5 PVE Servers. Running in the same issue, since good a year. It's only on VMs when snapshots. I read about an ext and LVM bug on Redhat bugzilla. A VM damaged the whole hostfilesystem. Nice.
To you have snapshots on this VM?