12-3inode耗尽故障处理

2016.6.25

12-3-1inode耗尽导致磁盘故障

  • 由于每个文件都必须有一个inode，因此有可能发生inode已经用光，但是硬盘还未存满的情况，这时，就无法再硬盘上创建新文件
  • 案例
  • 查找原因
    • /data/cache目录中存在数量非常多的小字节缓存文件，占用的bloc不多，但是占用了大量的inode
  • 解决方案
    • 删除/data/cache目录中的部分文件，释放出/data分区的一部分inode

12-3-2inode耗尽磁盘故障

  • 模拟实验思路

    1新建一个约32M大小的EXT4文件系统
    2编写一个测试程序，耗尽所有可用的inode
    3测试创建新文件
    4解决inode耗尽故障

通过/dev/sde4来分一个32M的主分区，然后格式化成ext4，挂载到/testinode下面，查看挂载信息。


[root@test2 ~]# fdisk /dev/sde

WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
         switch off the mode (command 'c') and change display units to
         sectors (command 'u').

Command (m for help): n
Command action
   e   extended
   p   primary partition (1-4)
p
Selected partition 4
First cylinder (787-2610, default 787):
Using default value 787
Last cylinder, +cylinders or +size{K,M,G} (787-2610, default 2610): +32m
Unsupported suffix: 'm'.
Supported: 10^N: KB (KiloByte), MB (MegaByte), GB (GigaByte)
            2^N: K  (KibiByte), M  (MebiByte), G  (GibiByte)
Last cylinder, +cylinders or +size{K,M,G} (787-2610, default 2610): +32M

Command (m for help): P

Disk /dev/sde: 21.5 GB, 21474836480 bytes
255 heads, 63 sectors/track, 2610 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xe4b1d138

   Device Boot      Start         End      Blocks   Id  System
/dev/sde1               1         262     2104483+  fd  Linux raid autodetect
/dev/sde2             263         524     2104515   fd  Linux raid autodetect
/dev/sde3             525         786     2104515   fd  Linux raid autodetect
/dev/sde4             787         791       40162+  83  Linux

Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
[root@test2 ~]# partprobe /dev/sde
sde   sde1  sde2  sde3  sde4
[root@test2 ~]# cd /
[root@test2 /]# mkdir /testinode
[root@test2 /]# mount /dev/sde4 /testinode
mount: you must specify the filesystem type
[root@test2 /]# mkfs
mkfs          mkfs.ext2     mkfs.ext4     mkfs.msdos
mkfs.cramfs   mkfs.ext3     mkfs.ext4dev  mkfs.vfat
[root@test2 /]# mkfs.ext4 /dev/sde4
mke2fs 1.41.12 (17-May-2010)
Filesystem label=
OS type: Linux
Block size=1024 (log=0)
Fragment size=1024 (log=0)
Stride=0 blocks, Stripe width=0 blocks
10040 inodes, 40160 blocks
2008 blocks (5.00%) reserved for the super user
First data block=1
Maximum filesystem blocks=41156608
5 block groups
8192 blocks per group, 8192 fragments per group
2008 inodes per group
Superblock backups stored on blocks:
    8193, 24577

Writing inode tables: done
Creating journal (4096 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 39 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.
[root@test2 /]# mount /dev/sde4 /testinode/
[root@test2 /]# df -hT | grep /dev/sde4
/dev/sde4                    ext4    38M  4.5M   32M  13% /testinode
[root@test2 /]#

查看新挂载硬盘的inodedf -i /testinode/

[root@test2 /]# df -i /testinode/
Filesystem     Inodes IUsed IFree IUse% Mounted on
/dev/sde4       10040    11 10029    1% /testinode
[root@test2 /]# df -i /dev/sde4
Filesystem     Inodes IUsed IFree IUse% Mounted on
/dev/sde4       10040    11 10029    1% /testinode
[root@test2 /]#

创建一个小程序来耗尽inode：

[root@test2 /]# cd /testinode/
[root@test2 testinode]# touch test
[root@test2 testinode]# vi test

#!/bin/bash
for n in $(seq 1 10040)
do
touch a_$n
done
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
:wq

这就是老师曾经说的用for循环来创建

[root@test2 testinode]# mv test test.sh
[root@test2 testinode]# sh test.sh
touch: cannot touch `a_10029': No space left on device
touch: cannot touch `a_10030': No space left on device
touch: cannot touch `a_10031': No space left on device
touch: cannot touch `a_10032': No space left on device
touch: cannot touch `a_10033': No space left on device
touch: cannot touch `a_10034': No space left on device
touch: cannot touch `a_10035': No space left on device
touch: cannot touch `a_10036': No space left on device
touch: cannot touch `a_10037': No space left on device
touch: cannot touch `a_10038': No space left on device
touch: cannot touch `a_10039': No space left on device
touch: cannot touch `a_10040': No space left on device
[root@test2 testinode]#

根据上面的信息能看出来Ifree：10029. 1-10028正好是10028，我觉得那1个inode可能是被test.sh给占用了，毕竟，你是先查看的inode，在你写好了脚本之后并没有再次进行查看。

再次查看sde4的inode：df -i /dev/sde4

[root@test2 testinode]# df -i /dev/sde4
Filesystem     Inodes IUsed IFree IUse% Mounted on
/dev/sde4       10040 10040     0  100% /testinode

df -h /dev/sde4
[root@test2 testinode]# df -h /dev/sde4
Filesystem      Size  Used Avail Use% Mounted on
/dev/sde4        38M  4.7M   32M  14% /testinode
[root@test2 testinode]#

使用for循环删除

[root@test2 testinode]# vi test.sh
#!/bin/bash
for n in $(seq 1 10040)
do
rm -rf a_$n
done
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
:wq

[root@test2 testinode]# sh test.sh  //运行程序删掉了所有以a开头的文件
[root@test2 testinode]# ls -l
total 13
drwx------ 2 root root 12288 Jun 25 17:20 lost+found
-rw-r--r-- 1 root root    57 Jun 25 17:35 test.sh
[root@test2 testinode]#

其实这里你可以再次查看一下inode的信息

results matching ""

    No results matching ""