使用逻辑卷管理灵活储存

Posted 2020-09-14

第五单元

使用逻辑卷管理灵活储存

一  lvm 定义

Lvm是建立在硬盘和分区之上的一个逻辑层，来提高磁盘分区管理的灵活性。将磁盘分区处理成物理卷（pv），再将物理卷变成一个整块的卷组（voilmegroup，vg）形成一个储存池。然后就可以在卷组随意创建lvm（逻辑卷组），并进一步在lvm上创建文件系统。

Lvm可拉伸也可压缩，当系统添加了新的磁盘，通过lvm，就不用将磁盘的文件移动到新的磁盘上以充分利用新的存储空间。而是直接扩展文件系统跨越磁盘即可

 

二 lvm 建立

如果出现以下问题：

[[email protected] ~]# pvcreate /dev/vdb1

  Physical volume /dev/vdb1 not found

  Can‘t open /dev/vdb1 exclusively.  Mounted filesystem?

说明/dev/md0没有停止

 

解决：[[email protected] ~]# mdadm -S /dev/md0

mdadm: stopped /dev/md0

 

 

1  fdisk    /dev/vdb            ###建立两个分区####

图

注意：在wq保存前，要先将分区类型选成lvm的类型

监控pv，vg，lvm的信息：watch   -n   1   ‘echo  ===pvmessaage=== ; pvs; echo ===vgmessage=== ; vgs; echo ===lvmessage=== ; lvs;  df  /weixindata’

2  pvcreate   /dev/vdb1                    ###将物理分区创建成物理卷（pv）###

3   vgcreate   vg0    /dev/vdb1           ##将物理卷创建成卷组，卷组名为vg0###

4  lvcreate   -L   300M   -n   lv0   vg0   ###在vg0上创建lvm ，-L 值大小为300M，-n指取名为lv0###

5  Mkfs.xfs    /dev/vg0/lv0                ###将lvm格式化###

6  mount   /dev/vg0/lv0   /weixindata/        ###将lvm挂载到/weixindata目录下

Lvm创建过程如下：

[[email protected] ~]# pvcreate /dev/vdb1

  Physical volume /dev/vdb1 not found

WARNING: linux_raid_member signature detected on /dev/vdb1 at offset 4096. Wipe it? [y/n] y

  Wiping linux_raid_member signature on /dev/vdb1.

WARNING: xfs signature detected on /dev/vdb1 at offset 0. Wipe it? [y/n] y

  Wiping xfs signature on /dev/vdb1.

  Physical volume "/dev/vdb1" successfully created

[[email protected] ~]# vgcreate vg0 /dev/vdb1

  Volume group "vg0" successfully created

[[email protected] ~]# lvcreate -L 300M -n lv0 vg0

  Logical volume "lv0" created

[[email protected] ~]# mkfs.xfs /dev/vg0/lv0

meta-data=/dev/vg0/lv0           isize=256    agcount=4, agsize=19200 blks

         =                       sectsz=512   attr=2, projid32bit=1

         =                       crc=0

data     =                       bsize=4096   blocks=76800, imaxpct=25

         =                       sunit=0      swidth=0 blks

naming   =version 2              bsize=4096   ascii-ci=0 ftype=0

log      =internal log           bsize=4096   blocks=853, version=2

         =                       sectsz=512   sunit=0 blks, lazy-count=1

realtime =none                   extsz=4096   blocks=0, rtextents=0

[[email protected] ~]# mount /dev/vg0/lv0 /weixindata/

 

 

 注： /dev/mapper/vg0-lv0 和  /dev/vg0/lv0都是指你所创建的lvm即lv0

如下：

[[email protected] ~]# ll /dev/mapper/vg0-lv0

lrwxrwxrwx. 1 root root 7 Apr 22 03:47 /dev/mapper/vg0-lv0 -> ../dm-0

[[email protected] ~]# ll /dev/vg0/lv0

lrwxrwxrwx. 1 root root 7 Apr 22 03:47 /dev/vg0/lv0 -> ../dm-0

[[email protected] ~]#

 

三 lvm拉伸

分两步：1 拉伸设备

       2 拉伸文件系统

Lvextend  -L  900M   /dev/vg0/lv0    ###拉伸lvm到900M###

xfs_growfs   /dev/vg0/lv0             ###拉伸文件系统###

过程如下：

[[email protected] ~]# lvextend -L 900M /dev/vg0/lv0

  Extending logical volume lv0 to 900.00 MiB

  Logical volume lv0 successfully resized

[[email protected] ~]# xfs_growfs /dev/vg0/lv0

meta-data=/dev/mapper/vg0-lv0    isize=256    agcount=4, agsize=19200 blks

         =                       sectsz=512   attr=2, projid32bit=1

         =                       crc=0

data     =                       bsize=4096   blocks=76800, imaxpct=25

         =                       sunit=0      swidth=0 blks

naming   =version 2              bsize=4096   ascii-ci=0 ftype=0

log      =internal               bsize=4096   blocks=853, version=2

         =                       sectsz=512   sunit=0 blks, lazy-count=1

realtime =none                   extsz=4096   blocks=0, rtextents=0

data blocks changed from 76800 to 230400

 

注意：如果要拉伸的大小太大，lvm不够时，只需扩大pv即可，pv不够就再加一块硬盘。

例：将大小为1024M的lvm拉伸到1500M，由于lvm不够，因此要扩大pv

Pvcreate   /dev/vdb2     ###增加pv###

vgextend   vg0   /dev/vdb2   ###增加vg0###

Lvextend   -L    1500M    /dev/vg0/lv0   ###拉伸lvm到1500M###

xfs_growfs     /dev/vg0/lv0     ###拉伸文件系统###

过程如下：

[[email protected] ~]# lvextend -L 1500M /dev/vg0/lv0

  Extending logical volume lv0 to 1.46 GiB

  Insufficient free space: 150 extents needed, but only 30 available

[[email protected] ~]# pvcreate /dev/vdb2

  Physical volume /dev/vdb2 not found

WARNING: linux_raid_member signature detected on /dev/vdb2 at offset 4096. Wipe it? [y/n] y

  Wiping linux_raid_member signature on /dev/vdb2.

  Physical volume "/dev/vdb2" successfully created

[[email protected] ~]# vgextend vg0 /dev/vdb2

  Volume group "vg0" successfully extended

[[email protected] ~]# lvextend -L 1500M /dev/vg0/lv0

  Extending logical volume lv0 to 1.46 GiB

  Logical volume lv0 successfully resized

[[email protected] ~]# xfs_growfs /dev/vg0/lv0

meta-data=/dev/mapper/vg0-lv0    isize=256    agcount=12, agsize=19200 blks

         =                       sectsz=512   attr=2, projid32bit=1

         =                       crc=0

data     =                       bsize=4096   blocks=230400, imaxpct=25

         =                       sunit=0      swidth=0 blks

naming   =version 2              bsize=4096   ascii-ci=0 ftype=0

log      =internal               bsize=4096   blocks=853, version=2

         =                       sectsz=512   sunit=0 blks, lazy-count=1

realtime =none                   extsz=4096   blocks=0, rtextents=0

data blocks changed from 230400 to 384000

 

 

 

四 lvm 压缩

注意：有的文件系统不能压缩，像xfs的文件系统就不能压缩，ext4文件系统可以压缩

所以如果是xfs的文件系统要先改变文件系统

Umount   /weixindata            ###卸载###

Mkfs.ext4   /dev/vg0/lv0         ###格式化###

Mount   /dev/vg0/lv0   /weixindata/   ###挂载###

过程如下：

[[email protected] ~]# umount /weixindata/

[[email protected] ~]# mkfs.ext4 /dev/vg0/lv0

mke2fs 1.42.9 (28-Dec-2013)

Filesystem label=

OS type: Linux

Block size=4096 (log=2)

Fragment size=4096 (log=2)

Stride=0 blocks, Stripe width=0 blocks

96000 inodes, 384000 blocks

19200 blocks (5.00%) reserved for the super user

First data block=0

Maximum filesystem blocks=394264576

12 block groups

32768 blocks per group, 32768 fragments per group

8000 inodes per group

Superblock backups stored on blocks:

32768, 98304, 163840, 229376, 294912

 

Allocating group tables: done                            

Writing inode tables: done                            

Creating journal (8192 blocks): done

Writing superblocks and filesystem accounting information: done

 

[[email protected] ~]# mount /dev/vg0/lv0 /weixindata/

开始压缩：

为了防止压缩过小将文件损坏，就要先卸载，然后扫描，如果压缩过小就会提示

umount    /weixindata/     ###卸载###

e2fsck   -f   /dev/vg0/lv0     ###扫描###

resize2fs   /dev/vg0/lv0   300M   ###压缩文件系统到300M###

如果没有提示，则说明压缩在允许范围内

mount    /dev/vg0/lv0     /weixindata/   ##挂载###

lvreduce   -L   300    /dev/vg0/lv0    ###压缩设备到300M###

 

过程如下：

[[email protected] ~]# umount /weixindata/

[[email protected] ~]# mkfs.ext4 /dev/vg0/lv0

mke2fs 1.42.9 (28-Dec-2013)

Filesystem label=

OS type: Linux

Block size=4096 (log=2)

Fragment size=4096 (log=2)

Stride=0 blocks, Stripe width=0 blocks

96000 inodes, 384000 blocks

19200 blocks (5.00%) reserved for the super user

First data block=0

Maximum filesystem blocks=394264576

12 block groups

32768 blocks per group, 32768 fragments per group

8000 inodes per group

Superblock backups stored on blocks:

32768, 98304, 163840, 229376, 294912

 

Allocating group tables: done                            

Writing inode tables: done                            

Creating journal (8192 blocks): done

Writing superblocks and filesystem accounting information: done

 

[[email protected] ~]# mount /dev/vg0/lv0 /weixindata/

 

 

五 移除物理卷

要从一个卷组中移除一个物理卷，首先要确认要移除的物理卷没有被任何逻辑卷正在使用，可通过“pvdisplay”命令查看

Pvdisplay    /dev/vdb2

如果total   PE 和Free  PE 相等，则没有被使用

如果正被使用，就需要将该物理卷的数据转移到别的物理卷，然后才能移除。

pvmove   /dev/vdb1   /dev/vdb2  ###将物理卷/dev/vdb1的数据移到/dev/vdb2###

vgreduce  vg0   /dev/vdb1        ###把vdb1从vg0里移走###

pvremove    /dev/vdb1           ###删除物理卷/dev/vdb1 ###

过程如下：

[email protected] ~]# pvmove /dev/vdb1 /dev/vdb2

  /dev/vdb1: Moved: 17.3%

  /dev/vdb1: Moved: 100.0%

[[email protected] ~]# vgreduce vg0 /dev/vdb1        

  Removed "/dev/vdb1" from volume group "vg0"

[[email protected] ~]# pvremove /dev/vdb1            

  Labels on physical volume "/dev/vdb1" successfully wiped

 

 

 

 

六 lvm 快照

目的 ：为保护原始数据

 lvcreate  -L  50M  -n  lv0backup  -s  /dev/vg0/lv0 ###-s指对/dev/vg0/lv0进行快照，-L指快照大小为50M，-n指快照名字为 lv0backup

mount   /dev/vg0/lv0backup    /weixindata/     ###将快照挂载到/weixindata/

 

例：在lvm挂载的目录/weixindata/ 下建立文件，然后对lvm进行快照，将其挂载到/weixindata/ 下，查看里面的文件，将文件全部删除，再删除快照，重新建立新的快照，再次查看/weixindata/ 下文件，发现文件还在，所以起到保护原始数据的作用。

过程如下：

[[email protected] ~]# touch  /weixindata/file{1..10}

[[email protected] ~]# cd /weixindata/

[[email protected] weixindata]# ls

file1   file2  file4  file6  file8  lost+found

file10  file3  file5  file7  file9

[[email protected] weixindata]# lvcreate -L 50M -n lv0backup -s /dev/vg0/lv0

  Rounding up size to full physical extent 52.00 MiB

  Logical volume "lv0backup" created

[[email protected] weixindata]# mount /dev/vg0/lv0backup /weixindata/

[[email protected] weixindata]# cd /weixindata/

[[email protected] weixindata]# ls

file1   file2  file4  file6  file8  lost+found

file10  file3  file5  file7  file9

[[email protected] weixindata]# rm -fr *

[[email protected] weixindata]# ls

[[email protected] weixindata]# cd

[[email protected] ~]# umount /weixindata/

[[email protected] ~]# lvremove /dev/vg0/lv0backup       ###删除快照###

Do you really want to remove active logical volume lv0backup? [y/n]: y

  Logical volume "lv0backup" successfully removed

[[email protected] ~]# lvcreate -L 50M -n lv0backup -s /dev/vg0/lv0  ###建立快照##

  Rounding up size to full physical extent 52.00 MiB

  Logical volume "lv0backup" created

[[email protected] ~]# mount /dev/vg0/lv0backup /weixindata/

[[email protected] ~]# cd /weixindata/

[[email protected] weixindata]# ls

file1   file2  file4  file6  file8  lost+found

file10  file3  file5  file7  file9

 

七 卸载

umount /weixindata/                ###卸载###

lvremove /dev/vg0/lvbackup        ###删除快照###

lvremove /dev/vg0/lv0            ###删除lv0###

vgremove vg0                     ###删除vg0###

pvremove /dev/vdb2              ###删除物理卷/dev/vdb2###