Manoj Namdev

8 months ago · 4 min. reading time · ~10 ·

Blogging
>
Manoj blog
>
Migrate RAID disks to LVM volumes

Migrate RAID disks to LVM volumes

Goal: Migrate data from existing Raid5 Array to lvm partitions 

Action: Current raid5 disk array is having 4 disks, we have to remove one by one disk from /dev/md0 raid5 array and use them for the lvm partitioning scheme. Once lvm created from the raid disks (/dev/sdb8) we’ll copy our data from existing raid array to lvm mount point.

Copy Raid5 data to newly created lvm partition(/dev/vg1/lv1) mount point /lv1

Once we are copied whole data from the raid5 to lvm, we’ll wipe out /raid5 devices and use them in newly created lvm.

First, we need to convert all raid devices from 'Linux raid autodetect' to 'Linux LVM' 

Then we have to create pv for all three devices

Then vgextend for the existing volume group vg1

Then lvextend for exiting lvm partition mounted on /lv1

Resize the filesystem type

Put entry inside /etc/fstab file

[root@node1 ~]# mdadm -D /dev/md0

[root.@nodel /dev/mdo: Version Creation Time Raid Level Array § Used Dev Size Raid Devices Total Devices Persistence LC Update Time State Active Devices rking Devices Failed Devices RISES SERIA Nel) W Layout Chunk Size Consistency Policy Eu uuIn Fvents Number Major 0 8 I 8 2 8 ) 8 t@nodel ~)§ ~1# mdadm -D /dev/md0 1 Fri Dec 16 01:23:30 2022 [RES] ERIC RTE) 1046528 ) ) Superblock (2.99 GiB 3.21 CB) (1022.00 MiB 1071.64 MB) 1s persistent Mon Dec 19 09:26:43 2022 clean 4 4 left-symmetric B34 jgeRAv iT nodel:0 (local to host nodel) A121¢741:8d67253eece0141d: 44144641 A RN Raidhevice State 21 0 active sy /dev/sdbh 22 I active /dev/sdbé A] 2 active sy JAS EVs oy) px! E] active sy /dev/sdb8

Check mount point and raid array name, check disk size as well.

[root@node1 ~]# df -h

[root@nodel ~]§ df -h |W EEY Size devimpts JI tmpfs JI tmpfs JI tmpfs JI /dev/mapper/centos LEH /dev/mdo ATS /dev/sdal LEAR [RS 280M [root@nodel ~)§ [ | 9.1M 195M NEE LEE LEE LEE LEE IRIS RRs EVAN] 280M ounted on /dev /dev/shm /run vs/ts/cqroup 1s JATRLSE /boot. /run/user/0

Mark commented on raid5 mount point in /etc/fstab so it won’t be problem in case we need reboot to update partition tables as we’ll be going to use existing disk into the lvm partition.

[root@node1 ~]# vim /etc/fstab

/dev/mapper/centos-swap swap BAtE¢) defaults

Mark fail any single device which you want to remove from the existing raid5 array (/dev/sdb8)

[root@node1 ~]# mdadm /dev/md0 --fail /dev/sdb8

[root@nodel ~]# mdadm /dev/md0 --fail /dev/sdb8 mdadm: set /dev/sdb8 faulty in /dev/md0O [root@nodel ~]§ [ |

Unmount raid5 mount point

[root@node1 ~]# umount /dev/md0

Remove failed disk from the array which was failed in last command /dev/sdb8

[root@node1 ~]# mdadm /dev/md0 --remove /dev/sdb8

[root@nodel ~]# mdadm /dev/md0 --remove /dev/sdb8 mdadm: hot. removed /dev/sdb8 from /dev/md0O [root@nodel ~]§ [ |

Check existing array should report failed/remove device from the raid5 array /dev/md0

[root@node1 ~]# mdadm -D /dev/md0

[root@nodel ~]1# mdadm -D /dev/md0O JAS E74 SON] Version : 1.2 Creation Time Fri Dec 16 01:23:30 2022 Raid Level raidh Array Siz 3139584 (2.99 GiB 3.21 GB) Used Dev Size 1046528 (1022.00 MiB 1071.64 MB) Raid Devic ] Total Devices E] Persistence Superblock is persistent Mon Dec 19 09:43:45 2022 clean, degraded Update Time State Active Devices 03 Working Devic FE) Failed Devices © 0 Spare Devices : 0 Layout Chunk Size left-symmetric RNA Consistency Policy : resyne Name : nodel:0 (local to host nodel) A121¢741:8d67253c10000141d: 44144641 [AVIS RII AV ISIPHSTee Major SERN RaidbDevice State 0 15} 21 0 active sync /dev/sdbh I! 15} Vy I! active sync /dev/sdb6 2 15} Vai 2 active sync /dev/ sdb 0 E] removed & 0 [root@nodel ~]§ [|

Resize existing raid5 as one disk has been removed from the array.

[root@node1 ~]# resize2fs -P /dev/md0

[root@nodel ~]1# resize2ts -pP /dev/md0 g ze2is 1.42.9 (28-Dec-2013) stimated minimum size of the filesystem: 26476 t@nodel ~)§

Check raid status it should not have any current process going on, and should show 1 removed device (you can see there is only 3 U’s inside the bracket [UUU_] 

[root@node1 ~]# cat /proc/mdstat

@nodel ~1# cat /proc/mdstat nalitiecs : [raidé] [raidb] [raid4d] md : active raid) sdbb [0] sdb7([2] sdb6[1] 3139584 blocks super 1.2 level 5, 512k chunk, algorithm 2 [4/3] [UUU |] unused devices: <none> [root@nodel ~)§

Check and update filesystem accordingly for the raid5 array

[root@node1 ~]# e2fsck -f /dev/md0

[root@nodel ~]& e2tsck -f /dev/md0o sek 1.42.9 (28-Dec-2013) Checking inodes, blocks, and sizes Checking directory structure Checking directory connectivity Checking reference counts : Checking group summary information /dev/md0: 16/196224 files (0.0% non-contiguous), 310167784896 blocks [root@nodel ~18

Shrink the filesystem (-M) and print (-p) progress of the shrinking so once shrinking completed, we can go ahead for the next operation.

[root@node1 ~]# resize2fs -p -M /dev/md0

ze2ts -p -M /dev/md0 [root@nodel ~1§ r i 3-Dec=2013) resize2ts 1.42.9 ( Resizing the filesystem on /dev/md0 to 26476 (4k) blocks. Begin pass 2 (max 16386) Relocating blocks ).9.9.9.0.:9.9.9.9.9.9.9.9.9.9.9.9.0.9.99.99990999.990999900990004 Begin pass 3 (max a) Scanning inode table ).9.9.9.0.:9.9.9.9.9.9.9.9.9.9.9.9.0.9.99.99990999.990999900990004 The filesystem on /dev/md0 is now 26476 blocks long. [root@nodel ~18

To check array-size show /dev/md9 detail from the below command and record array size.

[root@node1 ~]# mdadm -D /dev/md0

[root.@nodel JAS E74 SON] NZS EARN Creation Time Raid Level Array Size Used Dev Size Raid Devices Total Devices Persistence Update Time State |DISAVAREIexH EARS |DISAVAREIexH |DISAVAREIexH LAL IRIs) Working JERR RSs] Spare Layout Chunk Size Consistency Policy Name uuIn Fvents IS[FhsTe a J ERT 0 8 I 15} 2 8 - § ] | root. @nodel ~| 8 [1] ~1# mdadm -D /dev/md0O 1 Fri Dec 16 01:23:30 2022 raidh BARE 1046528 ] E] Superblock (2.99 GiB 3.21 GB) (1022.00 MiB 1071.64 MB) is persistent Mon Dec 19 09:55:58 2022 clean, degraded E] E] 0 0 left-symmetric RNA resync nodel:0 (local to host nodel) 4121¢741:8d67253c100c0141d: 44144641 a [IRiIete RaidDevice State 21 0 active sync /dev/sdbh 22 I active sync /dev/sdbé Vai 2 active sync /dev/ sdb 0 E] removed

Need to adjust /dev/md0 array, so copy array-size from above command

[root@node1 ~]# mdadm --grow /dev/md0 --array-size 3139584

Resize existing array once again to update the existing disk array size

[root@node1 ~]# resize2fs /dev/md0

@nodel AS r e2is 1.42. zing the {110s The filesystem on Jed /md0 [RSE Stet] HOYT blocks 1 size2ts /dev/mdO y Dec=-2013) SITeH [Bat t.@nodel ~ 8 [ | on /dev/md0 to 784896 (4k) blocks.

Now active device size reduced to 3 disk device array

[root@node1 ~]# mdadm -D /dev/md0

[root.@nodel /dev/mdo: NZS EARN Creation Time Raid Level Array Size Used Dev Size Raid Devices ESHER ER IARS Persistence Update Time State PISIVAREIeN Devic PISIVAREIeN PISIVAREIeN LAS RIS: LIS Sse] JERR RSs] Spare Layout Chunk Size Consistency Policy Name uuIn Fvents NIFH TSE J ERT 0 8 I 15} 2 15} - 0 ~1# mdadm -D /dev/md0O 1 Fri Dec 16 01:23:30 2022 raidh BARE 1046528 ] E] Superblock (2.99 GiB 3.21 GB) (1022.00 MiB 1071.64 MB) is persistent Mon Dec 19 09:58:36 2022 clean, degraded 3 3 0 0 left-symmetric RNA resync nodel:0 (local to host nodel) 4121¢741:8d67253c100c0141d: 44144641 al [IRiIete Raidhevice State al 0} active sync /dev/sdbh 22 I active sync /dev/sdbé Vai 2 active sync /dev/ sdb 0 E] removed

Reboot to update the partition table

[root@node1 ~]# reboot

Mount raid5 array and check data inside the raid5 mount point, we will go ahead and copy this data to lvm partition

[root@node1 ~]# mount /dev/md0 /raid5/

[root@node1 ~]# ls /raid5

[root@nodel ~1§ ls /raidb dirl dir? filel helloworld lost+found profile [RESIS N CI STS (SR IE IF |]

[root@node1 ~]# wipefs -a /dev/sdb8

[root@nodel ~]1# wipetfs -a /dev/sdb8 /dev/sdb8: 4 bytes were erased at offset 0x00001000 (linux raid member): fe de 2b a9 [root@nodel ~]§ [ |

Changed type of partition 'Linux raid autodetect' to 'Linux LVM'

[root@node1 ~]# fdisk /dev/sdb

Command (m for help): t Partition number (1,5-8, default 8): 8 Hex code (type I to list all codes): Se Changed type of partition 'Linux raid autodetect! to 'Linux LVM! Command (m for help): p Disk /dev/sdb: 16.1 GB, 16106127360 bytes, 31457280 sectors Units sectors of 1% 512 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes 1/0 size (minimum/optimal): 512 bytes / 512 bytes Disk label type: dos Disk identifier: Oxabd3676% Device Boot Ri¥tad 13313] Blocks Id System /dev/sdbl 2048 EARINY MEE] ERMAN) 5H Extended /dev/sdbh 4096 2101247 LEE RYEY fd Linux raid autodetect /dev/sdbé 2103296 LAE LEE RYEY fd Linux raid autodetect /dev/ sdb 4202496 6299647 1048576 fd Linux raid autodetect /dev/sdb8 6301696 8398847 LEE RYEY 8¢ Linux LVM

Create new lvm from the raid5 wiped disk and copy data from /raid5 array to newly created and mounted lvm partition /lv1

[root@node1 ~]# pvcreate /dev/sdb8

[root@node1 ~]# vgcreate vg1 /dev/sdb8

[root@node1 ~]# lvcreate -L +900M -n lv1 vg1

[root@node1 ~]# mkfs.xfs /dev/vg1/lv1

[root@node1 ~]# mkdir /lv1

[root@node1 ~]# mount /dev/vg1/lv1 /lv1

[root@node1 ~]# cp -r /raid5/* /lv1

[root@node1 ~]# ls /lv1

[root@nodel ~1# pvereate /dev/sdb8 Physical volume "/dev/sdb8" successfully created. [root@nodel ~]# vgcereate vgl /dev/sdbg Volume group "vgl" successfully created [root@nodel ~]§ lvcereate -1, 900M -n lvl vgl Logical volume "lvl" created. [root@nodel ~]1# mkis.xfs /dev/vgl/livl meta-data /dev/vgl/lvl isize H12 agcecount. 4, agsize 7600 blks K attr 2, projid32bit 1 finobt 0, sparse 0 data blocks 230400, imaxpcet 25 swidth 0 blks [GSLENITS| version 2 ascii-ci 0 ftype 1 RSLs] internal log blocks 85%, version 2 sunit 0 blks, lazy-count. 1 blocks 0, rtextents 0 realtime none [root@nodel ~]# mkdir /1lvl [root@nodel ~]§ mount. /dev/vgl/lvl [lvl [root@nodel ~]§ df -h Filesystem Size Used Avail Use? Mounted on devtmpts LEE 0 1.4c¢ OE VASTLY [BUTS LEE 0 1.4c¢ /dev/shm [BUTS 1.4G¢ 8.7M 1.46 /run tmpfs |S 0 1.46 FAVA EaSitle) /dev/mapper/centos-root. 13G 1.66 IRIS 1s /dev/sdal 1014M 195M 820M /boot. tmpfs 280M [ORS T0I | /run/user/0 Jas tea ste 2.96 9.1M 2.8G /raidb /dev/mapper/vgl-1vl CENA] 33M 865M JAR! [root@nodel ~1# cp -r /raidb/* [lvl [root@nodel ~1§ 1s /1vl dirl dir? filel helloworld lostifound profile [root@nodel ~18

We can see raid5 data is on the newly created lvm partition, now if we are okay and confident that we have copied all the data from the raid5 and accommodate in the newly created lvm partition then it would be safe to remove raid array from the system.

Be careful: please double check data and its size on the newly created lvm partition and then only do the below operation.

Remove all the disks from the existing array.

[root@node1 ~]# mdadm /dev/md0 --fail /dev/sdb7

[root@node1 ~]# mdadm /dev/md0 --remove /dev/sdb7

[root@nodel ~]# mdadm /dev/md0 --fail /dev/sdb7 mdadm: set /dev/sdb7 faulty in /dev/md0 [root@nodel ~]# mdadm /dev/md0 --remove /dev/sdb7 mdadm: hot. removed /dev/sdb7 from /dev/md0O [root@nodel ~]# mdadm -D /dev/md0

Check the raid5 details and its disks which is currently participated into the raid5 array.

[root.@nodel /dev/mdo: NZS EARN Creation Time Raid Level Array Size Used Dev Size Raid Devic ESHER ER IARS Persistence Update Time State EARS EARS EARS PISIVARErS) Ja avs Working JERR RSs] Spare Layout Chunk Size Consistency Policy Name uuIn Fvents NIFH TSE J ERT 0 8 I 15} - 0 = 9) ~1# mdadm -D /dev/md0O 1 Fri Dec 16 01:23:30 2022 raidh BARE 1046528 ] 2 Superblock (2.99 GiB 3.21 GB) (1022.00 MiB 1071.64 MB) is persistent 19 11:21:20 2022 | AN BB AY] Mon Dec clean, left-symmetric RNA resync nodel:0 (local to host nodel) 4121¢741:8d67253c100c0141d: 44144641 RE [IRiIete Raidhevice State al 0} active sync /dev/sdbh Vy I! active sync /dev/sdb6 0 Yi removed 0 E] removed

[root@node1 ~]# e2fsck -f /dev/md0

[root@nodel ~]1# mdadm --stop /dev/md0 mdadm: stopped /dev/md0O [root@nodel ~]1# mdadm --zero-superblock /dev/sdbb [root@nodel ~]1# mdadm --zero-superblock /dev/sdb6 [root@nodel ~]1# mdadm --zero-superblock /dev/sdb? [root@nodel ~]§ [root@nodel ~]# reboot [ |

Note: Raid5 does not allow us to remove any disk completely from raid5, having read error during the filesystem check

We have to fail and remove all the existing disks and stop raid5 array and set zero superblock for all the remaining raid devices.

[root@node1 ~]# mdadm /dev/md0 --fail /dev/sdb5

[root@node1 ~]# mdadm /dev/md0 --fail /dev/sdb6

[root@node1 ~]# mdadm /dev/md0 --remove /dev/sdb5

[root@node1 ~]# mdadm /dev/md0 --remove /dev/sdb6

As we have copied data from raid5 array, we can stop and wiped out all the disks and use them for the newly created lvm volume. We’ll extend volume group and then logical volume.

First of all, unmount /dev/md0 and set zero superblock for all the faulty and removed disks.

[root@node1 ~]# umount /dev/md0

[root@node1 ~]# mdadm --stop /dev/md0

[root@node1 ~]# mdadm --zero-superblock /dev/sdb5

[root@node1 ~]# mdadm --zero-superblock /dev/sdb6

[root@node1 ~]# mdadm --zero-superblock /dev/sdb7

[root@node1 ~]# reboot

Command (m for help): t Partition number (1,5-8, default 8): 7 Hex code (type I to list all codes): Se Changed type of partition "Linux raid autodetect to Command (m for help): t Partition number (1,5-8, default 8): 6 Hex code (type I to list all codes): Se Changed type of partition "Linux raid autodetect to Command (m for help): t Partition number (1,5-8, default 8): 9 Hex code (type I to list all codes): Se Changed type of partition "Linux raid autodetect to Command (m for help): p Disk /dev/sdb: 16.1 GB, 16106127360 bytes, 31457280 Units sectors of 1% 512 512 bytes Sector size (logical/physical): 512 bytes / 512 byte 1/0 size (minimum/optimal): 512 bytes / 512 bytes Disk label type: dos Disk identifier: Oxabd3676% Device Boot Ri¥tad 13313] Blocks /dev/sdbl 2048 EIR EVV] ERM ELAN /dev/sdbh EIETS Fav 1048576 /dev/sdbé 2103296 EVAR EY LEE RYEY /dev/ sdb 4202496 6299647 1048576 /dev/sdb8 6301696 8398847 1048576 AR ARINIPY AR ARINIPY AR ARINIPY sectors X81 H Ge Ge Ge Ge System LVM' LVM' LVM' Extended Linux Linux Linux Linux LVM LVM LVM LVM

Changed type of partition 'Linux raid autodetect' to 'Linux LVM'

[root@node1 ~]# fdisk /dev/sdb

[rootdnodel ~]§ partprobe [root@nodel ~]1# pvereate /dev/sdbb /dev/sdb6 /dev/sdbi Physical volume "/dev/sdbH" stully created. Physical volume "/dev/sdbé" stully created. Physical volume "/dev/sdbi" stully created. [root@nodel ~]§ [root@nodel ~]# vgextend vgl /dev/sdbd /dev/sdbe /dev/sdbl Volume group "vgl" successfully extended [root@nodel ~]§ [root@nodel ~]1# vagdisplay vgl --- Volume group --- ASI BY 18 (Zeal System 1D Format. [RG Metadata Areas 4 Metadata Sequence Noo 3 NSRP YSTEe TE read/write VG Status [SEPIA NS MAX LV 0 Cur LV I! Open LV 0 Max PV 0 Cur pV 4 Act. PV El] VCR PAS 3.98 GiB [HEPA 4.00 MiB ESI INO) 1020 Alloc PE / Size 22% / 900.00 MiB Free PE / Size 795 / <3.11 GiB \CINeIeR EY] 2Fxw6C-KANT-KPmN=-1{Qm | -F 1 pW-mhCF-qdJmFD

[root@node1 ~]# partprobe

[root@node1 ~]# pvcreate /dev/sdb5 /dev/sdb6 /dev/sdb7

[root@node1 ~]# vgextend vg1 /dev/sdb5 /dev/sdb6 /dev/sdb7

[root@node1 ~]# vgdisplay

[root@nodel ~]1# lvextend -1, +36 /dev/vgl/lvl Size of logical volume vgl/1lvl changed from 900.00 MiB (225 extents) to <3.88 GiB (993 extents). Logical volume vgl/lvl successfully resized. [root@nodel ~]§ [root@nodel ~]§ xis growts /dev/vgl/livl xts growfs: /dev/vgl/lvl is not a mounted XFS filesystem [root@nodel ~]§

[root@node1 ~]# lvextend -L +3G /dev/vg1/lv1

[root@node1 ~]# xfs_growfs /dev/vg1/lv1

[root@nodel ~]# vad Volume group [ETSY NE ZN System 1D Ens [Re Metadata Areas 1 Metadata Sequence Noo 4 VG Access [REVO Tae Rn NERY [REP EEO 0 Cur LV 1 Open LV 0 EET 0 [Sa 1 Fa 1 A'¢ [Rds] 3.98 1B PE Size ERI FIP 1020 [RT AA Tes LERNER BTR) [TUT AE IP 21 7 108 EHS PERT 2¥xwébe KANT KPmN Om) FlpW mhCF qddmbD [root@nodel ~]1# lvdisplay /dev/vgl/lvl Logical volume [ARON Estate! LV Name 181 IH MW: 7S xBpd 11rd ZJBS elym Koyk MUUSqg/ Tate LV Write Acc LV Creation host, time nodel, FRVETET LV Status EARS Eh [sa 0 IV Size PETIT Current LE ELE ETL RTTe 1 Allocation inherit Read ahead sectors ETtIES currently set to [ICV Block device 2

[root@node1 ~]# lvdisplay

[root@nodel ~]# df h RE8aY JH JH JH JH JH JIE] PEI) er/centon root EVE 8 HE) i] [root@nodel ~]# mount /dev/vgl/lvl /1vl [root@nodel ~]# df h Avail JH JH JH JH 116 ERC] PEI) Used F1 lesyntem Size Used Avail devtmpt [RTS 1.46 3 JH 01.46 3 1.4G 6.6M 1.46 STR [TH 01.46 /dev/mapper/centos root 136 1.66 116 LEVEE 1014M 195M 820M SIR PEI) 0 280M fdev/mapper/vgl lvl EER] EEL BET] [rootBnodel ~]1# xfs growfs /dev/vgl/lvl neta data /dev/mapper/vgl lvl [EPI HEPA ! SET buize 4096 ERUEE aN ELIE REET buize 4096 IE) internal buize 4096 sectsz 512 SENSE TER TTI extaz 4096 data blocks changed from 230400 to 1016832 Mounted on /dev /dev/shm /run EE Ey ran a /boot. /run/user/0 % Mounted on PE rae am RET li /boot. rss Jase ageount 4, agmize 51600 blks EE REFIT LRT THe [LEER blocks 230400, imax swidth 0 blks TA I TR a) blocks 855, version 2 sunit 0 blks, lazy count 1 blocks 0, rtextents

[root@node1 ~]# df -h

[root@node1 ~]# mount /dev/vg1/lv1 /lv1

[root@node1 ~]# xfs_growfs /dev/vg1/lv1

[root@nodel ~]1# pvs IY VG Fmt. Attr PSize |RSS /dev/sda? centos lvm? a-- <14.00qg 0 /dev/sdb8 val lvm? a-- 1020.00m 120.00m [root@nodel ~1# lvdisplay /dev/vgl/lvl --- Logical volume --- LV Path /dev/vagl/lvl LV Name [al NEES (Zeal LV UuID GeaHuK=-AWnz-oWi i -RVeN-Ri Fv-{ Tt.S-Qreby0 LV Write Access read/write LV Creation host, time nodel, 2022-12-19 16:48:39 +0530 LV status available LES STs I! LV Size 900.00 MiB Current. LE 225 Segments 1 Allocation inherit Read ahead sectors ELIAS) - currently set to ERM Block device 25312 [root@nodel ~1# df -h | grep lvl /dev/mapper/vgl-1vl CENA] 33M 865M 4% [lvl [root@nodel ~]§ [ |

[root@node1 ~]# df -h

[root@nodel ~]§ lvcereate - 100M RRPAS snapshot. --name Logical volume “snap lvl" created. [root@nodel ~1§ lvs LV VG Attr JIS IP2S! Pool Origin Data% [aeTeld centos -—wi-ao---- <12.50g swap centos —wi-ao---- J ISYOLS | v1 NZS owi=-aos--- 900.00m snap lvl vgl swi-a-s--- 100.00m [Ra 0.00 [root@nodel ~1# df -h | grep lvl /dev/mapper/vgl-1vl CENA] 33M 865M 4% [lvl [root@nodel ~]§ [ | snap [SANE Move lvl /dev/vgl/ivl Log Cpy%Sync Convert

LVM snapshot and restore data of lv1 logical volume

[root@node1 ~]# pvs

[root@node1 ~]# lvdisplay /dev/vg1/lv1

[root@node1 ~]# df -h | grep lv1

[root@nodel ~)§& 1s [lvl anaconda-ks.ctg dirl dir? filedata libaudit.so.l libaudit.so.1.0.0 [root@nodel ~1# df -h | grep lvl /dev/mapper/vgl-1vl CENA] 33M 865M 4% [lvl [root@nodel ~18

Create lvm snapshot for the disk /dev/vg1/lv1 to backup lv1 data

[root@node1 ~]# lvcreate --size 100M --snapshot --name snap_lv1 /dev/vg1/lv1

[root@nodel ~]1# dmsctup status vgl-snap lvl-cow: 0 204800 linear vgl-lvl-real: 0 1843200 linear vgl-snap lvl: 0 1843200 snapshot. 72/204800 16 vgl-1vl: 0 1843200 snapshot-origin centos-swap: 0 3145728 linear centos-root: 0 26206208 linear [root@nodel ~]§

We have some files on /lv1 mount point which is currently available in the lvm snapshot.

root@nodel ~]¢# df -h | grep lvl /dev/mapper/vgl-1vl CENA] 33M 865M 4% [lvl root@nodel ~]# root@nodel ~]# umount. /1vl root@nodel ~]# root@nodel ~]1# mount. /dev/vgl/snap lvl /mnt root@nodel ~]# root@nodel ~]# 1s anaconda-ks.ctg ZFS (1).tar.gz zis-release.el’ d.noarch.rpm zisrpms root@nodel ~]# root@nodel ~]# umount. /mnt. root@nodel ~]# root@nodel ~]1§ mount. /dev/vgl/lvl /lvl root@nodel ~]# root@nodel ~]# 1s anaconda-ks.ctg ZFS (1).tar.gz zis-release.el’ d.noarch.rpm zisrpms root@nodel ~]# JANN JANN tar tar SFA SFA

We can use dmsetup status to check snapshot metadata sectors.

[root@nodel ~]# umount. /1vl [root@nodel ~]# mount. /dev/vgl/lvl /mnt [root@nodel ~1# df -h | grep lvl /dev/mapper/vgl-1vl CENA] 33M 865M 4% /mnt. [root@nodel ~]§ [root@nodel ~]# tar -cvzi snap lvl.tag.gz /mnt/ tar: Removing leading */' from member names /mnt./ /mnt./libaudit.so.l /mnt./libaudit.so.1.0.0 /mnt./anaconda-ks.cfqg VATA NRE EL /mnt./dirl/ /mnt./dirl/anaconda-ks.ciqg EWEN /mnt./dir?2/anaconda-ks.ctqg [root@nodel ~1§ clear [root@nodel ~18

Check data on the snapshot after unmounting actual mount point before mounting snapshot volume.

[root@node1 ~]# df -h | grep lv1

[root@node1 ~]# umount /lv1

[root@node1 ~]# mount /dev/vg1/snap_lv1 /mnt

[root@node1 ~]# umount /mnt

[root@node1 ~]# mount /dev/vg1/lv1 /lv1

[root@nodel ~]§ 1s -1 total 56 —rw------- «1 root root 1227 Dec 16 00:40 anaconda-ks.ctqg -rw-r--r--. 1 root root 49311 Dec 19 18:23 snap lvl.tag.qgz drwxr-xr-x. 3 root. root 35 Dec 19 16:10 zisrpms [root@nodel ~1§ 1s

[root@node1 ~]# umount /lv1

[root@node1 ~]# mount /dev/vg1/lv1 /mnt

[root@node1 ~]# df -h | grep lv1

[root@node1 ~]# tar -cvzf snap_lv1.tag.gz /mnt/

t@nodel ~1# lveonvert --mergesnapshot. /dev/vgl/snap lvl Delaying merge since origin is open. Merging of snapshot vgl/snap lvl will occur on next activation of vgl/lvl. -oot@nodel ~|§ @nodel ~]¢# lvchange --refresh vgl/lvl t@nodel ~]§ @nodel ~]§ df -h Size Used Avail Us Mounted on devtmpis |S 0 1.46 /dev tmpfs |S 0 1.46 /dev/shm tmpfs 1.46 8.6M 1.4G /run tmpfs |S 0 1.46 /sys/ts/cgroup /dev/mapper/cent 13G 1.86 IRIS 1s /dev/sdal 1014 195M 820M /boot. tmpfs 280M [ORS T0I | /run/user/0 /dev/mapper/vgl-1vi 897M RII ETRY FALLS [root@nodel ~1§ 1s /mnt anaconda-ks.ctfqg [root@nodel ~1§ lvs WV VG [XEN Lsize Pp root. cent -wi-ao---- <12.50qg swap centos -—wi-—¢ --- J RRSTOLe | [Ra (7601 Owi =: ERIM IONT0) 1) [root@nodel ~)§ [ | rl dir? filedata libaudit.so.1l libaudit.so.1.0.0 Origin Data% Metas Mc ¢ Log Cpy%Sync Convert

[root@node1 ~]# ls

rmIBS.png

Restore snapshot to the original logical volume.

[root@node1 ~]#  lvconvert --mergesnapshot /dev/vg1/snap_lv1

 We can refresh the logical volume for it to reactivate using the latest metadata using “lvchange” 

[root@node1 ~]# lvchange --refresh vg1/lv1

[root@node1 ~]# ls /mnt

Now lvsnapshot is removed and merged with the original logical volume.

SsucM.png

#LVM #mount #LEE #JH #Linux
Science and Technology
Comments

You may be interested in these jobs

  • CareerPaths

    SAP FICO FM Consultant

    Found in: Foundit IN A2 - 6 days ago

    CareerPaths Delhi, India Full time

    1.Expertise on SAP Finance (FI) & Funds Management (FM) module · On demand support to Business Process and Operational Divisions Teams · Accountable for Solution issues escalated to the Support team ...

  • FIRON CONSULTING

    Relationship Manager

    Found in: beBee S2 IN - 3 days ago

    FIRON CONSULTING Pune, India Full time

    Strengthen distribution by identifying potential leaders · Support and manage the BAs in Recruitment & Training of Advisors. · Manage Advisor career progressions and create a pool of Premier Advisor ...

  • PSR FASHIONS AND LIFESTYLES PVT LTD

    Customer Care Executive

    Found in: Emprego IN C2 - 3 days ago

    PSR FASHIONS AND LIFESTYLES PVT LTD Coimbatore, India

    We are looking for a passionate and skilled Customer Care Executive to join our team at The Nesavu, a vibrant and growing kidswear brand. · In this role, you will be the first point of contact for our ...