In the Fedora 17 machine, a Hard drive which was mounted on /mnt/c suddenly became read-only.
The fstab contents showed the UUID of the ext4 drives. With the help of the command “blkid” I could determine that /dev/sdc1 was mounted to /mnt/c (linux-finding-using-uuids-to-update-fstab)

After umounting and mounting with several commands nothing helped. Here is what i did without success:
I tried “mount -o remount, rw /dev/hdc1″ which stated that i should give the filesystem type as a parameter …
I tried  ”umount -l /dev/hdc1″ to force umounting. (umount-device-is-busy-fehler-beim-unmounten-eines-laufwerks)
I killed every process which was using /mnt/c (lsof | grep “mnt/c”)

After restarting the machine the drive was writable again but several minutes later it was back in a read-only state and even worse i could not read some directories (I/O Error). I found out, that maybe the drive was (re)mounted with the option”errors=remount-ro" (unix.stackexchange.com) But the output of the command on my machine was “Errors: Continue” so this were very bad news because the filesystem or the drive must be broken.

The Command “fsck.ext4 -cDfty -C 0 /dev/sd” showed that there were no errors (howto-repair-broken-ext4-partitions). I used it in a GNU Screen so that this time consuming command was not attached to a shell. (screen).

After accessing the drive with the scp command the second time the filesystem was broken again (problem after restart identically):

[nils@gres ~]$ /usr/local/bin/bakupSql.sh
xx.xx                                         100%   23MB  11.6MB/s   00:02    
[nils@xxx ~]$  /usr/local/bin/bakupSql.sh
scp: /mnt/c/bak/xx/xx.xx: Read-only file system
scp: /mnt/c/bak/xx/xx.xx: truncate: Read-only file system
[1]+  Done                    /usr/local/bin/bakupSql.sh

Several commands showed, that this drive definitely had some issues:

[root@xx ~]# umount /mnt/c
umount: /mnt/c: not mounted
[root@xx ~]# mount /dev/sdc1 /mnt/c
mount: /dev/sdc1: can't read superblock
[root@xx ~]# umount /mnt/c
umount: /mnt/c: not mounted 
[root@xx ~]# fsck.ext4 -cDfty -C 0 /dev/sdc1 
e2fsck 1.42.3 (14-May-2012)
fsck.ext4: Ein Block konnte nicht in einem Zug gelesen werden beim Versuch, /dev/sdc1 zu ?ffnen
K?nnte es eine Partion der L?nge Null sein?