SUSE Linux case file system read-only, force uninstall

SUSE Linux case file system read-only, force uninstall

Problem background
In the morning, we found that all backups failed yesterday, and the error was/Oracle_bk file system read-only.

Handling process
Bi-db :~ # Lsb_release-
LSB Version: core-2.0-noarch: core-3.0-noarch: core-2.0-x86_64: core-3.0-x86_64: desktop-3.1-amd64: desktop-3.1-noarch: graphics-2.0-amd64: graphics-2.0-noarch: graphics-3.1-amd64: graphics-3.1-noarch
Distributor ID: SUSE LINUX
Description: SUSE Linux Enterprise Server 10 (x86_64)
Release: 10
Codename: n/
Bi-db :~ # Cat/etc/SuSE-release
SUSE Linux Enterprise Server 10 (x86_64)
VERSION = 10
PATCHLEVEL = 2

Bi-db :~ # Df-h

Filesystem Size Used Avail Use % Mounted on
/Dev/cciss/c0d0p2 104G 94G 4.4G 96%/
Udev 16G 1.4 M 16G 1%/dev
/Dev/mapper/biemcvg-biemclv
738G 698G 2.0G 100%/oracle
/Dev/cciss/c0d1p5 1.4 T 1.3 T 58G 96%/oracle_bk

Go to the/oracle_bk directory and touch a file and report an error, so I plan to unmount the file system and re-mount it.

Bi-db :~ # Umount/oracle_bk
Umount:/oracle_bk: device is busy

Umount:/oracle_bk: device is busy

If the device is busy, you can check who is using it.

Bi-db :~ # Lsof | grep/oracle_bk

Bi-db :~ # Fuser-k-m-v/oracle_bk

It's strange that no one is using it and there are no processes.

If you run the command again, an error is returned.

Bi-db :~ # Umount/oracle_bk
Umount:/oracle_bk: device is busy
Umount:/oracle_bk: device is busy

In another way, execute the uninstallation task (see), and the result is successful. In fact, if not, you can also force uninstall (-f parameter)
Bi-db :~ # Umount-l/oracle_bk
Bi-db :~ # Df-h
Filesystem Size Used Avail Use % Mounted on
/Dev/cciss/c0d0p2 104G 94G 4.4G 96%/
Udev 16G 1.4 M 16G 1%/dev
/Dev/mapper/biemcvg-biemclv
738G 698G 2.0G 100%/oracle

Bi-db :~ # Mount/oracle_bk
Bi-db :~ # Df-h
Filesystem Size Used Avail Use % Mounted on
/Dev/cciss/c0d0p2 104G 94G 4.4G 96%/
Udev 16G 1.4 M 16G 1%/dev
/Dev/mapper/biemcvg-biemclv
738G 698G 2.0G 100%/oracle
/Dev/cciss/c0d1p5 1.4 T 1.3 T 58G 96%/oracle_bk
Bi-db :~ # Cd/oracle_bk
Bi-db:/oracle_bk # touch aaa
Bi-db:/oracle_bk # ls-al aaa *
-Rw-r -- 1 root 0 Jul 24 aaa
Bi-db:/oracle_bk # rm aaa

Important umount parameters:

-A: delete all file systems in fstab.
-F enforces umount, mainly for inaccessible NFS. This should be risky. You should use it with caution. After all, it's really tough.

-L in the lazy mode, delete the file system level first. When the file system is not busy, clear all connections.

Order of use of parameters:

Umount/aaa

Umount-l/aaa

Umount-f/aaa

Summarize troubleshooting ideas

The problem of file system read-only is still quite common. As a veteran, you 'd better be aware of the handling ideas before you can be surprised.

1) Make sure that all users have exited the mount point directory and exited the applications that occupy the directory, especially their own

2) run the fuser or lsof command to check whether there are dead processes in use and kill them.

3) umount corresponds to the file system. For the priority of the methods used, refer to the content above this article.

File System Repair case in SUSE 11