I can't start any service I got Read-Only error message.

[evolver]

Hello All ,

O got this error message today on my server.

For example I was try restart to exim ;


[root@gateway ~]# service exim restart
Shutting down exim: /etc/init.d/exim: line 40: kill: (3107) - No such process

rm: cannot remove `/var/lock/subsys/exim': Read-only file system
Starting exim:
touch: cannot touch `/var/lock/subsys/exim': Read-only file system

Please help me...

I can't create a file too :( :( How can I solve this problem ?

NOTE: This problem is begins after server reboot. (I think hard reboot)

[sullise]

just go into the /var/lock/subsys folder and manually delete the lock file.


killall -9 exim
rm -rf /var/lock/subsys/exim
service exim restart

see if that works.

[nobaloney]

Generally when a file system is marked read-only it's because it was found to have hardware problems. Then your OS marks the system as read-only to protect it from further damage.

The first step would be a local boot, looking for error messages flashing by on the screen.

If you're running Linux you may have to run fsck, but that could do even more damge.

If you don't already have backups, make some now (you'll probably have to make them directly to a different drive/partition).

Jeff

[chartmusic]

Originally posted by jlasman
Generally when a file system is marked read-only it's because it was found to have hardware problems. Then your OS marks the system as read-only to protect it from further damage.

The first step would be a local boot, looking for error messages flashing by on the screen.

If you're running Linux you may have to run fsck, but that could do even more damge.

If you don't already have backups, make some now (you'll probably have to make them directly to a different drive/partition).

Jeff

Not local to the system so can't local boot.
Where to look in log files for informaiton please?
Also is it safe/useful to do reboot?

[nobaloney]

See my other post. Rebooting may help you; but your server may just decide to take hours to boot (if it's trying to run fsck itself) or not boot at all.

What kind of information are you looking for in the log files? General information is in /var/log/messages, but i don't know what you're trying to look for.

Usually if you've got a hardware problem you're going to end up needing local hands.

Jeff

[chartmusic]

Originally posted by jlasman
See my other post. Rebooting may help you; but your server may just decide to take hours to boot (if it's trying to run fsck itself) or not boot at all.

What kind of information are you looking for in the log files? General information is in /var/log/messages, but i don't know what you're trying to look for.

Usually if you've got a hardware problem you're going to end up needing local hands.

Jeff
Was looking for some information to show what files or areas where corrupt. Basically something to see what could be done to fix it. The whole system seems to be available at first but in read-only state. Within a hour or so it all shut down and I couldn't access the system at all re-motely - No SSH. Was hoping to do a backup or something useful.
In the event it looked bad. But a new second disk was installed with OS re-installed and the original 'bad' disk still attached. When the 'bad' disk was mounted it had all the important stuff on it in good condition anyway.
Looking back it seems a bit of an over-reaction for the system to go into a read-only state.

Certainly learnt some things anyway. :)

[nobaloney]

You can always rewrite the linux kernel if you'd like :) .

If the system thinks the drive is corrupt it goes into read-only mode to protect the data.

You may think that's over-reaction.

I don't.

Jeff