Linux disk management and file system (e)

knowledge structure:

  • Disk basis

  • Detect and confirm the new disk

  • Planning hard disk partition

  • Create a file system

  • Mounting and unmounting file systems



(A) basic disk


Disk: Disk (Disk) memory means for storing data using magnetic recording technology. Computer disk is the main storage medium that can store large amounts of binary data after power failure and data loss can be maintained. Early computer disk using a floppy disk (soft disk, a floppy disk for short), is now commonly used disk hard disk (hard disk, referred to as hard disk).

Mechanical hard drive:

Screenshot micro-channel _20190819004545.png

The physical structure of the hard drive

  • Disk: a plurality of hard disks, each disk surface 2

  • Head: the head on each side a

Screenshot micro-channel _20190819004545.png

Hard disk data structures

  • Sector: the disk is divided into a plurality of segment areas, each sector 512 bytes of data storage

  • Track: the same concentric circles of different radii of the disk

  • Cylindrical: cylindrical surface composed of the same radius disk nowhere

Screenshot micro-channel _20190819004545.png

Screenshot micro-channel _20190819005102.png

Disk capacity is calculated:

  • Storage capacity = number of heads × tracks (cylinders) × number of sectors per track Number of bytes per sector ×

  • FIG 3 is a disk disk heads 6, 7 cylinders (tracks per disc 7) of the magnetic disk, FIG. 3 has 12 sectors per track, so this disk capacity:

The storage capacity of 6 * 7 * 12 * 512 = 258048

  • Number of sectors per track is the same as said old disk, the outer ring small density, the density of the inner ring, the amount of data stored per revolution is the same. Densities of hard disk data are consistent, so the longer the circumferential length of the track, the more sectors, the greater the amount of data stored.

Disk read response time:

  • Seek: time to start moving the head from the track where the data needed, the shorter the seek time, I / O operations faster average seek time is currently generally disk 3-15ms, generally about 10ms.

  • Rotational latency: the disk rotation data requested sectors where write head moves downward time required, depending on the disk rotational speed delay. Ordinary hard disk are generally 7200rpm, slow 5400rpm.

  • Data transfer time: time of transmission of the requested data needed to complete.

Summary about: From the above indicators, the most important thing, or that we should be most concerned with only two: seek time; rotational latency.

The time required to read and write information on a disk can be decomposed into: seek time, latency, transmission time. In order to improve the efficiency of disk transfer, the software should consider reducing the seek time and latency.

Block / cluster:

Outline

磁盘块/簇(虚拟出来的)。 块是操作系统中最小的逻辑存储单位。操作系统与磁盘打交道的最小单位是磁盘块。

通俗的来讲,在Windows下如NTFS等文件系统中叫做簇;在Linux下如Ext4等文件系统中叫做块(block)。每个簇或者块可以包括2、4、8、16、32、64…2的n次方个扇区。

为什么存在磁盘块?

读取方便:由于扇区的数量比较小,数目众多在寻址时比较困难,所以操作系统就将相邻的扇区组合在一起,形成一个块,再对块进行整体的操作。

分离对底层的依赖:操作系统忽略对底层物理存储结构的设计。通过虚拟出来磁盘块的概念,在系统中认为块是最小的单位。

page

操作系统经常与内存和硬盘这两种存储设备进行通信,类似于“块”的概念,都需要一种虚拟的基本单位。所以,与内存操作,是虚拟一个页的概念来作为最小单位。与硬盘打交道,就是以块为最小单位。

扇区、块/簇、page的关系

  • 扇区: 硬盘的最小读写单元

  • 块/簇: 是操作系统针对硬盘读写的最小单元

  • page: 是内存与操作系统之间操作的最小单元。

扇区 <= 块/簇 <= page

Screenshot micro-channel _20190819004545.png

磁盘的接口类型:

IDE(并口)

SATA(串口)

SCSI

MBR与磁盘分区表示

主引导记录(MBR:Master Boot Record)

  • MBR位于硬盘第一个物理扇区处

  • MBR中包含硬盘的主引导程序和硬盘分区表

  • 分区表有4个主分区,每个分区记录占16个字节

Screenshot micro-channel _20190819004545.png

  • 硬盘中的主分区数目只有4个

  • 因此主分区和扩展分区的序号也就限制在1~4

  • 扩展分区在分为逻辑分区

  • 逻辑分区的序号将始终从5开始

Linux中将硬盘,分区等设备均表示为文件

Screenshot micro-channel _20190819005102.png

文件系统类型

XFS文件系统

  • 存放文件和目录数据的分区

  • 高性能的日志型文件系统

  • Centos7系统中默认使用的文件系统

SWAP,交换文件系统

  • 为Linux系统建立交换分区

Linux支持的其他文件系统类型

  • FAT16,FAT32,NTFS

  • EXT4,JFS



(二)检测并确认新硬盘

fdisk:查看或管理磁盘分区

fdisk -l [磁盘设备]
或
fdisk [磁盘设备]

Screenshot micro-channel _20190819005102.png

  • Device:分区的设备文件名称。 

  • Boot:是否是引导分区。若是,则有“*”标识。

  • Start:该分区在硬盘中的起始位置(柱面数)。

  • End:该分区在硬盘中的结束位置(柱面数)。

  • Blocks:分区的大小,以 Blocks(块)为单位,默认的块大小为 1024 字节。

  • Id:分区对应的系统 ID 号。83 表示 Linux 中的 XFS 分区或 EXT4 分区、8e 表示 LVM 逻 辑卷。

  • System:分区类型。

(三)规划硬盘中的分区

交互模式中的常用的命令

Screenshot micro-channel _20190819004545.png

m  帮助信息

Screenshot micro-channel _20190819005102.png

p  显示分区列表

Screenshot micro-channel _20190819024120.png

n  创建分区

Screenshot micro-channel _20190819024444.png

d  删除分区

Screenshot micro-channel _20190819025123.png

t  转换分区类型 id转换

Screenshot micro-channel _20190819024934.png

w  保存修改

q  退出

(四)创建文件系统

创建文件系统(格式化)-mkfs

mkfs -t 文件系统类型 分区设备
或
mkfs.xfs 分区设备

Screenshot micro-channel _20190819031037.png

Screenshot micro-channel _20190819031054.png

创建交换文件系统(格式化)-mkswap

mkswap 分区设备
swapon 分区设备
swapoff 分区设备
/proc/meminfo | grep SwapTotal 查看交换分区大小

Screenshot micro-channel _20190819031902.png

(五)挂载,卸载文件系统-mount umount

挂载文件系统,IOS镜像到指定文件夹

mount [-t 类型] 存储设备 挂载点目录
mount -o loop ISO镜像文件 挂载点目录

Screenshot micro-channel _20190819033636.png

卸载文件系统

umount 设备文件或者挂载点


Screenshot micro-channel _20190819033849.png

设置文件系统自动挂载

/etc/fstab配置文件

包含需要开机后自动挂载的文件系统记录

vim /etc/fstab

微信截图_20190819034332.png

/dev/sdb2:设备名称 

/opt/mnt1:挂载点 

xfs:文件系统类型 

defaults:功能权限  

0 0 not open without opening the intermediate disaster recovery check sequence starting with the distinction between the tab key or the spacebar


Refresh mount

mount -a mount refresh (auto-mount configuration file)

View disk usage

df [-hT]

"-H" option to display more readable unit of capacity

Corresponding to the type of the file system "-T" option is used to display

微信截图_20190819035136.png


After the words:

Disk and file management system is crucial in Linux, it is a basis for the entry of Linux! ! !

By slightly this article hoping to help you

thanks for reading! ! !


Guess you like

Origin blog.51cto.com/14080162/2430591