本文参考：https://blog.51cto.com/oldboy/612351编著

#模拟他的场景，防火墙关闭，SELinux关闭

安装Apache服务

[root@localhost /]# yum -y install httpd    #yum安装Apache服务

[root@localhost /]# service httpd restart   #重启apache服务

[root@localhost /]# grep -i   "^#Cu" /etc/httpd/conf/httpd.conf

#CustomLog logs/access_log common

#CustomLog logs/referer_log referer

#CustomLog logs/agent_log agent

#CustomLog logs/referer_log referer

#CustomLog logs/agent_log agent

[root@localhost /]# sed -i   "s@#CustomLog logs/access_log common@CustomLog /app/logs/access_log common@g"   /etc/httpd/conf/httpd.conf    #修改一下日志的存放路径

[root@localhost /]# grep -i "^#Cu" /etc/httpd/conf/httpd.conf           

#CustomLog /app/logs/access_log common

#CustomLog logs/referer_log referer

#CustomLog logs/agent_log agent

 

模拟创建一个小磁盘

[root@localhost /]# dd if\=/dev/mapper/VolGroup-lv_root of\=/dev/sdb1   bs\=8k count\=10

10+0 records in

10+0 records out

81920 bytes (82 kB) copied, 0.000201325   s, 407 MB/s

#dd模拟创建一个磁盘，if输入，of输出，bs：blocks大小，count：block的数量

 

[root@localhost /]# mkfs.ext4 /dev/sdb1   #格式化分区

mke2fs 1.41.12 (17-May-2010)

/dev/sdb1 is not a block special device.

Proceed anyway? (y,n) y

Filesystem label=

OS type: Linux

Block size=1024 (log=0)

Fragment size=1024 (log=0)

Stride=0 blocks, Stripe width=0 blocks

16 inodes, 80 blocks

4 blocks (5.00%) reserved for the super   user

First data block=1

1 block group

8192 blocks per group, 8192 fragments per   group

16 inodes per group

 

Writing inode tables: done                           

 

Filesystem too small for a journal

Writing superblocks and filesystem   accounting information: done

 

This filesystem will be automatically   checked every 28 mounts or

180 days, whichever comes first.  Use tune2fs -c or -i to override.

 

#挂载分区

[root@localhost /]# mkdir -p /app/logs/

[root@localhost /]# mount /dev/sdb1 /app/logs/      #这样挂载不成功，需要特使挂载，方法在下

mount: /dev/sdb1 is not a block device (maybe try `-o loop'?)

[root@localhost /]# mount -o loop /dev/sdb1 /app/logs/

[root@localhost /]# df -Th

Filesystem                   Type   Size    Used Avail Use% Mounted on

/dev/mapper/VolGroup-lv_root ext4    18G    990M   16G   6% /

tmpfs                        tmpfs  1.9G       0  1.9G   0% /dev/shm

/dev/sda1                    ext4   485M     33M  427M   8% /boot

/dev/sdb1                      ext4    73K   14K     55K  21% /app/logs

#一个循环访问的脚本，1.sh

#!/usr/bin/bash

for n in `seq 1000000`    #循环执行多少次

        do

        curl 12.1.1.13   &>/dev/null    #把访问这个网站的结果不管 对错全部放到/dev/null

        done 

[root@localhost /]# sh /tmp/1.sh &     #把这个脚本放到后台执行

[root@localhost /]# df -Th

Filesystem                   Type   Size    Used Avail Use% Mounted on

/dev/mapper/VolGroup-lv_root ext4    18G    995M   16G   6% /

tmpfs                        tmpfs  1.9G       0  1.9G   0% /dev/shm

/dev/sda1                    ext4   485M     33M  427M   8% /boot

/dev/sdb1                    ext4    73K     68K  1.0K  99% /app/logs

 

#解决方法一：删除日志文件，重启服务

[root@localhost /]# rm -rf /app/logs/access_log       #删除网站访问的日志文件

[root@localhost /]# df -Th

Filesystem                   Type   Size    Used Avail Use% Mounted on

/dev/mapper/VolGroup-lv_root ext4    18G    995M   16G   6% /

tmpfs                        tmpfs  1.9G       0  1.9G   0% /dev/shm

/dev/sda1                    ext4     485M   33M  427M     8% /boot

/dev/sdb1                    ext4    73K     68K  1.0K  99% /app/logs

#把日志文件删除之后发现磁盘还是满的，这时候是因为删除文件的原理是：硬链接数（i_linke）为0，进程占用（i_count）为0，这个时候虽然硬链接数为0，。但是进程占用数不为0，所以这个文件还时存在的

[root@localhost /]#/etc/init.d/httpd restart    #重启服务

Stopping httpd:                                              [  OK  ]

Starting httpd: httpd: Could not reliably   determine the server's fully qualified domain name, using 12.1.1.13 for   ServerName

                                                             [  OK  ]

[root@localhost /]# df -Th

Filesystem                   Type   Size    Used Avail Use% Mounted on

/dev/mapper/VolGroup-lv_root ext4    18G    995M   16G   6% /

tmpfs                        tmpfs  1.9G       0  1.9G   0% /dev/shm

/dev/sda1                    ext4   485M     33M  427M   8% /boot

/dev/sdb1                    ext4    73K     15K   54K  22% /app/logs

#把这个进程关闭在重启，然后进程占用为0，所以磁盘空间下来了。

 

#解决办法2，直接清空配置文件而不删除

[root@localhost /]# echo " " > /app/logs/access_log    #清空文件内容

[root@localhost /]# df -Th

Filesystem                   Type   Size    Used Avail Use% Mounted on

/dev/mapper/VolGroup-lv_root ext4    18G    996M   16G   6% /

tmpfs                        tmpfs  1.9G       0  1.9G     0% /dev/shm

/dev/sda1                    ext4   485M     33M  427M   8% /boot

/dev/sdb1                    ext4    73K     16K   53K  24% /app/logs


报错的解决办法：

#如果报一下错误，说明模拟的分区磁盘过小，不足以创建文件系统，我们只需要在dd指令的bs参数后面加上单位，给他大一点就行  

[root@localhost /]# mkfs.ext4  /dev/sdc1   

mke2fs 1.41.12 (17-May-2010)

/dev/sdc1 is not a block special device.

Proceed anyway? (y,n) y

mkfs.ext4: Device size reported to be   zero.  Invalid partition specified, or

          partition table wasn't reread after running fdisk, due to

          a modified partition being busy and in use.  You may need to reboot

          to re-read your partition table.