Equipment Management Under Linux

First, the device overview

1.1 Equipment Identity

Devices are in the form of a file after the access system

1.2 Device Name

SATA / SAS / USB	/dev/sda, /dev/sdb	s = SATA, d = disk, a, b represents the first few
HERE	/dev/hd0, /dev/hd1	h=hard
VIRTIO-BLOCK	/dev/vda, /dev/vdb	v=virtio
M2(SSD)	/dev/nvme0, /dev/nvme2	nvme representative Machine
SD / MMC / EMMC (card)	/dev/mmcblk0, /dev/mmcblk1	mmc representative Machine
CD-ROM	/dev/cdrom, /dev/sr0, /dev/sr1

1.3 Device View

command	content
fdisk -l	View disk partition (in the case of real systems)
lsblk	View device usage
blkid	View device usage and device id (id not find the device can not be mounted)
df	View equipment is being mounted system
cat /proc/partitions	View system identification device

Add a new SATA hard disk in the system

Second, the device mounts

The system has a device id device that can be used by system
conversion parameters from the read-only mount into a read-write
unloading equipment will display the target when the device is busy occupying
Note: The above are temporary mount equipment , you want to permanent mount devices need to edit the device to mount the policy file / etc / fstab
file format of the content to
the device mount point to mount the file system type parameter is detected whether the backup

can be mountused to view the file system commands type

command mount -a again read the contents of / etc / fstab in
will cause the system fails to restart, you need to follow the prompts to enter the root password in the user interface, comment out the line error, reboot the system when editing equipment policy file error

Third, find a device file

findCommand can be used to find the file device
common parameters are shown in Table

-name	file name
-user	File owner
-group	File has set
-type	file type
-perm	file permission
-exec	Excuting an order
-maxdepth	Find the deepest
-mindepth	Find the most superficial
-The	or
-a	and
-not	Invert Selection

Fourth, the device partition

Partition the device can more easily manage disk and recover data
partitioning:

Types of	Digit	Partition table size	The number of partitions support	A single partition size
Traditional partitioning (MBR)	32	64	4 primary partition, all partitions 16	2.2TB
UEFI(GPT)	64	128	Without limiting the theory, win the next 128	8ZB

4.1 MBR partitioning

Primary partition	Recording information of the master partition table partitions can be used directly
Extended partition	Master partition table recording partition, not directly, but logical partitions container
Logical partitions	Extended partitions on partitioning

When no extended partition, logical partition will not present, occupies the entire hard disk is preferably divided at the extended partition

4.2 MBR partitioning method

fdisk /dev/sdaCommon command parameters are as follows

parameter	content
m	help
d	Delete partition
l	List all partitions
n	New Partition
p	Display Partition
t	Change the partition type
w	Save Selection
q	drop out

When the device is busy or if other circumstances lead to added after the file has not been generated newly added partition, you need to execute the following command:

partprobe     ##同步分区表

Or the command:

partx  -d  /dev/sda   ##清理分区表
partx  -a  /dev/sda   ##重新加载分区表

Add partition or partitions of the

MBR partition can be divided into only three primary partition and an extended partition

Conversion 4.3 MBR partitioning into a GPT mode

Delete all of the above-mentioned manner MBR partition divided

Five, swap partition

Role 5.1 swap partition

All program data is in memory, when memory usage exceeds the limit, the system will not operate properly at runtime. To make the system more stable, as a part of the space division of a memory buffer on the hard drive - if the memory usage exceeds the limit, the kernel will idle memory data stored in the swap, when the program needs the data in the swap area, then the kernel the data swap partition back to the process of memory processing.

5.2 swap partition size recommendations

memory size	recommended swap partition size	When turn HIBERNATE swap partition size recommendations
The following 2GiB	Twice as much physical memory	Three times the physical memory
2-8GiB	等于物理内存	物理内存两倍
8-64GiB	至少4GiB	物理内存的1.5倍
64GiB以上	至少4GiB	HIBERNATE不开启

5.3 临时创建swap分区

新建一个分区
更改分区类型
同步分区表
格式胡以及激活分区
在激活分区时指定优先级，优先级的范围为0-32767，数字越大优先级越高

5.4 永久创建swap分区

需要更改/etc/fstab中的内容
再次激活分区并查看分区

5.5 删除swap分区

六、磁盘配额

6.1 磁盘配额的作用

指定用户能够写入指定设备的最大额度

6.2 磁盘配额的临时设定方法

新建一个分区
格式化新加的分区
设定指定用户wang的配额参数为20M
上图中各部分内容的含义

Filesystem	设备名称
blocks	用户已经创建的数据
soft	软限
hard	硬限
inodes	用户已经创建的文件个数
soft	软限
hard	硬限

6.2 磁盘配额的永久设定方法

编辑文件/etc/fstab

6.3 关闭磁盘配额

编辑文件/etc/fstab，将写入的配额参数去掉
关闭配额并进行测试

七、逻辑卷（LVM）

一些需要了解的名词

名词	解释
pv（物理卷）	被处理过的物理分区
pe（物理扩展）	设定的存储最小单元，LVM是由整数个pe组成的
vg（物理卷组）	捆绑pv到一个组中
lv（逻辑卷）	分配最终的使用设备

7.1 lvm设备建立

建立物理分区并设定分区类型为lvm
对物理卷，物理卷组，逻辑卷以及挂载点/qqdata进行监控
创建pv，vg并设定pe大小为2M，即设定好以后的lvm大小是2的整数倍
lvcreate -L 100M -n wanglv wangvg语句中-L 指定lvm大小，-n指定名称
在监控界面看到的情况

7.2 当vg中容量充足时lvm设备的拉伸

Stretching the device
file system is stretched. May use the device or the mount point rhel7 stretching in the file system can only add mount point rhel8

7.3 When the stretching device lvm insufficient capacity in vg

Plus a new type of partition lvm
monitoring interface to see the results of tensile
Note: xfs file system does not support the reduction can only be stretched, if the beginning of the design may need to be considered to reduce the need to use ext file format
using a command to change the file system

mkfs.ext4 /dev/wangvg/wanglv

Reduce the size of the
file device but the device has not been reduced down

Delete 7.4 lvm equipment

Deletion step:
1, umount /qqdata/unloading apparatus
2, lvremove /dev/wangvg/wanglvis removed LV
. 3, vgremove wangvgremove the VG
. 4, pvremove /dev/sda2remove the PV
. 5, the partition is removed