EXT2-fs error

Orinoprabel · 03-10-2007, 09:17 AM

Hi,

I have a Plesk 8.0.1 Dell PE 1850 server running with Linux CentOS 3 (2.4.21-27.0.2.ELsmp) and I have re-mounted my ../tmp on ext2 (Device: /dev/loop0 Mode: rw,nosuid,noexec) and all is running fine for almost two years now, suddenly (no update or upgrades done) I get this error messages in my logs:

kernel: EXT2-fs error (device loop(7,0)): ext2_new_block: Free blocks count corrupted for block group 3

And right after the error message a couple of customers on the phone

with the same error complaints: (“database error: [Got error 28 from table handler]”) on there MySQL databases. (these disappeared or was solved with repair table on phpMyAdmin)

1) What does the kernel error message mean

2) How can I fix this

3) Why does the error message corrupt some MySQL database tables

Thx in advance!

indig0kid · 03-10-2007, 12:26 PM

Try to use some of FS fixing tools, u can even load some LiveCD distro for suck kind of problem, or just try to use:

#fsck.ext2

hope this will help

rch1231 · 03-10-2007, 01:16 PM

Your problem comes from the fact that the /tmp directory is not really a partition with a files system but a file on the server and it is corrupted. Bring the server up in single user mode and unmount /var/tmp and /tmp and run the following as root:

fsck -y /usr/tmpDSK

then reboot and all will be well.

Orinoprabel · 03-11-2007, 03:29 AM

Thx! guy's I am going to work on it

Orinoprabel · 03-13-2007, 07:37 AM

Quote:

Your problem comes from the fact that the /tmp directory is not really a partition with a files system but a file on the server and it is corrupted. Bring the server up in single user mode and unmount /var/tmp and /tmp and run the following as root:

fsck -y /usr/tmpDSK

then reboot and all will be well.

Thank you very much, this was the exact problem, and it it solved now, thx to you!