[cfarm-users] gcc21 /scratch/6 device remounted ro after qemu-sparc64 executable run

[Aymeric]

On 22/07/2017 00:12, Pierre Muller via cfarm-users wrote:
> Hi all,
>
>  
>
> mount output:
> muller at gcc21:~/pas/test$ mount
> /dev/sdd on /scratch/6 type ext4 (rw,noatime,errors=remount-ro)
> while:
> muller at gcc21:~/pas/test$ cat /proc/mounts
> /dev/sdd /scratch/6 ext4 ro,noatime,errors=remount-ro,data=ordered 0 0
>
> Is it possible to remount /dev/sdd in rw on that machine,
> would that require a reboot?
>
> Sorry for the trouble...
>
Hi all,

I give a look, there are a lots of errors message in dmesg about the drive :
> [6994268.115054] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
> [6994268.115056] sd 6:0:0:0: [sdd] CDB:
> [6994268.115058] Read(10): 28 00 c5 41 11 28 00 00 08 00
> [6994268.115066] end_request: I/O error, dev sdd, sector 3309375784
> [6994268.116028] EXT4-fs warning (device sdd):
> __ext4_read_dirblock:674: error reading directory block (ino
> 103422206, block 0)
> [6994268.116788] sd 6:0:0:0: [sdd] 

I was able to umount the drive but not to remount neither to fsck so I'm
rebooting and hopping that everything will be fixed by a reboot.. else
we will need someone with physical access to investigate this issue
(probably a dead drive and I suppose not before Monday).

Regards,

-- 
Aymeric