記一次linux pv missing故障
阿新 • • 發佈：2018-12-05
現象描述：
linux一個邏輯卷lv_home丟失，無法正常掛載，之前掛載在/home目錄下:
[root@VM001 ~]# cat /etc/fstab
.......
/dev/VolGroup/lv_home /home ext4 defaults 0 0檢視檔案系統掛載情況，發現lv_home並沒有掛載：
[root@VM001 ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup-lv_root
20 嘗試手動掛載發現找不到lv_home這個邏輯卷：
[root@VM0001 ~]# mount /dev/VolGroup/lv_
lv_root lv_swap
[root@VM001 ~]# cd /dev/VolGroup/
[root@VM000000875 VolGroup]# ls
lv_root lv_swap
[root@VM001 VolGroup]# ll
total 0 檢視邏輯卷狀態，發現處於異常狀態(NOT available)：
[root@VM001 VolGroup]# lvdisplay
.......
--- Logical volume ---
LV Path /dev/VolGroup/lv_home
LV Name lv_home
VG Name VolGroup
LV UUID 4 [[email protected] VolGroup]# lvscan
ACTIVE '/dev/VolGroup/lv_root' [20.00 GiB] inherit
ACTIVE '/dev/VolGroup/lv_swap' [5.00 GiB] inherit
inactive '/dev/VolGroup/lv_home' [1.17 TiB] inherit問題排查：
執行vgdisplay -v命令有2個pv missing資訊，此處為關鍵報錯資訊：
[[email protected] ~]# vgdisplay -v
Using volume group(s) on command line.
Finding all volume groups.
Wiping cache of LVM-capable devices
There are 2 physical volumes missing.
There are 2 physical volumes missing.
--- Volume group ---
VG Name VolGroup
System ID
Format lvm2
Metadata Areas 4
...........檢視pv狀態：
[[email protected] ~]# pvs
PV VG Fmt Attr PSize PFree
/dev/xvda2 VolGroup lvm2 a-m 49.51g 0
/dev/xvda3 VolGroup lvm2 a-m 973.99g 0
/dev/xvde VolGroup lvm2 a-- 200.00g 0
/dev/xvdf1 VolGroup lvm2 a-- 1023.99g 1023.99g檢視pvs引數說明：
# man pvs
.......
NOTES
The pv_attr bits are:
1 (d)uplicate, (a)llocatable, (u)sed
2 e(x)ported
3 (m)issing可以確認/dev/xvda2和/dev/xvda3兩個物理卷狀態為missing(a-m).
問題解決：
用vgextend - -restoremissing命令成功恢復pv到正常狀態：
man vgextend有如下說明：
# man vgextend
.......
DESCRIPTION
vgextend adds one or more PVs to a VG. This increases the space available for LVs in the VG.
Also, PVs that have gone missing and then returned, e.g. due to a transient device failure, can be added back to the VG without re-initializing them
(see --restoremissing).
.......
- -restoremissing
Add a PV back into a VG after the PV was missing and then returned, e.g. due to a transient failure. The PV is not reinitialized
# 在PV丟失後再將PV新增到VG中，然後返回，例如由於瞬時故障。沒有重新初始化PV。命令格式：
# vgextend VG PV執行命令：
# vgextend --restoremissing VolGroup /dev/xvda2
# vgextend --restoremissing VolGroup /dev/xvda2執行完成以後檢視pv狀態，已經恢復正常：
[[email protected] ~]# pvs
PV VG Fmt Attr PSize PFree
/dev/xvda2 VolGroup lvm2 a-- 49.51g 0
/dev/xvda3 VolGroup lvm2 a-- 973.99g 0
/dev/xvde VolGroup lvm2 a-- 200.00g 0
/dev/xvdf1 VolGroup lvm2 a-- 1023.99g 1023.99g再檢視lv狀態，也恢復正常，能夠正常掛載並進行擴容：
[root@VM001 ~]# vgdisplay -v
Using volume group(s) on command line.
Finding all volume groups.
--- Volume group ---
VG Name VolGroup
System ID
............
--- Logical volume ---
LV Path /dev/VolGroup/lv_home
LV Name lv_home
VG Name VolGroup
LV UUID 4rdajn-62q9-RX5T-QorD-vLsh-hMEv-zxT1Qj
LV Write Access read/write
LV Creation host, time localhost.localdomain, 2016-03-17 09:19:33 +0800
LV Status available
# open 1
LV Size 2.15 TiB
Current LE 562815
Segments 4
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:2 手動執行掛載，此時lv_home邏輯卷掛載恢復正常：
[root@VM001 ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup-lv_root
20G 7.3G 12G 40% /
tmpfs 3.9G 224K 3.9G 1% /dev/shm
/dev/xvda1 477M 124M 328M 28% /boot
/dev/mapper/VolGroup-lv_home
2.2T 1.2T 876G 58% /home網上看到的其他類似問題：
案例1：
執行lvextend -l +100%FREE /dev/VolGroup00/LogVol00的時候，出現瞭如下錯誤:
Device still marked missing because of allocated data on it, remove volumes and consider bgreduce --removemissing.在查閱許多資料之後，發現可以使用restoremissing 引數來恢復
[root@Caron ~]# vgextend --restoremissing VolGroup00 /dev/sdh1再次執行lvextend命令的時候，提示成功。重新resize一下lvm大小：
[root@Caron ~]# resize2fs /dev/VolGroup00/LogVol00
[root@Caron ~]# df -h | more
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00 936G 612G 283G 69% /
devtmpfs 126G 0 126G 0% /dev
tmpfs 126G 0 126G 0% /dev/shm
tmpfs 126G 4.1G 122G 4% /run
tmpfs 126G 0 126G 0% /sys/fs/cgroup
/dev/sda1 283M 100M 164M 38% /boot
tmpfs 126G 4.1G 122G 4% /run/netns
tmpfs 26G 0 26G 0% /run/user/0案例2:
1.PV丟失的故障處理，手工建立邏輯卷失敗，報錯顯示有PV丟失：
[root@node29archive]# lvcreate -L 1G -n testlv vgcloud
Cannotchange VG vgcloud while PVs are missing.
Considervgreduce --removemissing.2.進一步檢視：
[root@node29archive]# pvs
PV VG Fmt Attr PSize PFree
/dev/mapper/mpathbp1 vgcloud lvm2 a-- 1.82t 89.40g
/dev/mapper/mpathbp2 vgcloud lvm2 a-- 1.82t 91.14g
/dev/mapper/mpathbp3 vgcloud lvm2 a-- 1.82t 30.14g
/dev/mapper/mpathbp4 vgcloud lvm2 a-- 1.64t 30.00g
/dev/mapper/mpathcp1 vgcloud lvm2 a-m 1.82t 84.14g
/dev/mapper/mpathcp2 vgcloud lvm2 a-m 1.82t 117.14g
/dev/mapper/mpathcp3 vgcloud lvm2 a-m 1.82t 1.33t
/dev/mapper/mpathcp4 vgcloud lvm2 a-- 1.81t 1.81t
[root@node29archive]# vgs
VG #PV #LV #SN Attr VSize VFree
vgcloud 8 126 0 wz-pn- 14.36t 3.58t其中發現有三個邏輯卷處於missing的狀態
3.經過查詢資料得知可以使用vgextend –restoremissing操作解決
[[email protected]]# vgextend --restoremissing vgcloud /dev/mapper/mpathcp1
Volume group "vgcloud" successfullyextended
[[email protected]]# vgextend --restoremissing vgcloud /dev/mapper/mpathcp2
Volume group "vgcloud" successfullyextended
[[email protected]]# vgextend --restoremissing vgcloud /dev/mapper/mpathcp3
Volume group "vgcloud" successfullyextended4.再看已經沒有missing狀態標示了,手工建立邏輯卷
[root@node29archive]# lvcreate -L 1G -n testlv vgcloud
Logical volume "testlv" created