[linux-lvm] Problem with mount QNAP disks in Linux

[linux-lvm] Problem with mount QNAP disks in Linux

[Daniel Łaskowski]

Hello,

I'm trying mount three disks from QNAP TS-469L in Linux (Fedora 25).
System correctly recognised disks and RAID5, but I have problem with 
activate LVM with data.

In QNAP QTS system it was look this way:
Storage Pool 1          5,44 TB
├─DataVol1 (System)     5,38 TB         (Free Size: ~300,00 GB)
└─DataVol2              509,46 GB       (Free Size: ~400,00 GB)

In Fedora it looks that disks and RAID are recognised OK.

$ lsblk
NAME            MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
sdb               8:16   0   2,7T  0 disk
├─sdb4            8:20   0 517,7M  0 part
│ └─md123         9:123  0 448,1M  0 raid1
├─sdb2            8:18   0 517,7M  0 part
│ └─md127         9:127  0 517,7M  0 raid1
├─sdb5            8:21   0     8G  0 part
│ └─md125         9:125  0   6,9G  0 raid1
├─sdb3            8:19   0   2,7T  0 part
│ └─md124         9:124  0   5,5T  0 raid5
│   └─vg1-lv544 253:3    0    20G  0 lvm
└─sdb1            8:17   0 517,7M  0 part
   └─md126         9:126  0 517,7M  0 raid1
sdc               8:32   0   2,7T  0 disk
├─sdc2            8:34   0 517,7M  0 part
│ └─md127         9:127  0 517,7M  0 raid1
├─sdc5            8:37   0     8G  0 part
│ └─md125         9:125  0   6,9G  0 raid1
├─sdc3            8:35   0   2,7T  0 part
│ └─md124         9:124  0   5,5T  0 raid5
│   └─vg1-lv544 253:3    0    20G  0 lvm
├─sdc1            8:33   0 517,7M  0 part
│ └─md126         9:126  0 517,7M  0 raid1
└─sdc4            8:36   0 517,7M  0 part
   └─md123         9:123  0 448,1M  0 raid1
sda               8:0    0   2,7T  0 disk
├─sda4            8:4    0 517,7M  0 part
│ └─md123         9:123  0 448,1M  0 raid1
├─sda2            8:2    0 517,7M  0 part
│ └─md127         9:127  0 517,7M  0 raid1
├─sda5            8:5    0     8G  0 part
│ └─md125         9:125  0   6,9G  0 raid1
├─sda3            8:3    0   2,7T  0 part
│ └─md124         9:124  0   5,5T  0 raid5
│   └─vg1-lv544 253:3    0    20G  0 lvm
└─sda1            8:1    0 517,7M  0 part
   └─md126         9:126  0 517,7M  0 raid1

$ cat /proc/mdstat
Personalities : [raid1] [raid6] [raid5] [raid4]
md123 : active (auto-read-only) raid1 sda4[2] sdc4[1] sdb4[0]
       458880 blocks super 1.0 [24/3] [UUU_____________________]
       bitmap: 1/1 pages [4KB], 65536KB chunk

md124 : active (auto-read-only) raid5 sdb3[0] sdc3[1] sda3[2]
       5840623232 blocks super 1.0 level 5, 64k chunk, algorithm 2 [3/3] 
[UUU]

md125 : active (auto-read-only) raid1 sda5[2] sdc5[1] sdb5[0]
       7168000 blocks super 1.0 [3/3] [UUU]
       bitmap: 0/1 pages [0KB], 65536KB chunk

md126 : active (auto-read-only) raid1 sda1[2] sdc1[1] sdb1[0]
       530112 blocks super 1.0 [24/3] [UUU_____________________]
       bitmap: 1/1 pages [4KB], 65536KB chunk

md127 : active raid1 sda2[2](S) sdb2[0] sdc2[1]
       530112 blocks super 1.0 [2/2] [UU]
       bitmap: 0/1 pages [0KB], 65536KB chunk

$ sudo pvs
   PV             VG     Fmt  Attr PSize   PFree
   /dev/md124     vg1    lvm2 a--    5,44t    0

$ sudo vgs
   VG     #PV #LV #SN Attr   VSize   VFree
   vg1      1   4   0 wz--n-   5,44t    0

$ sudo lvs
   LV    VG     Attr       LSize   Pool Origin Data% Meta%  Move Log 
Cpy%Sync Convert
   lv1   vg1    Vwi---tz--   5,42t tp1
   lv2   vg1    Vwi---tz-- 512,00g tp1
   lv544 vg1    -wi-a----- 20,00g
   tp1   vg1    twi---tz-- 5,40t

$ sudo lvscan
   ACTIVE            '/dev/vg1/lv544' [20,00 GiB] inherit
   inactive          '/dev/vg1/tp1' [5,40 TiB] inherit
   inactive          '/dev/vg1/lv1' [5,42 TiB] inherit
   inactive          '/dev/vg1/lv2' [512,00 GiB] inherit

If I understand correctly:
/dev/vg1/tp1 - it is QNAP "Storage Pool 1" (with real size 5,40 TiB)
/dev/vg1/lv1 - it is QNAP "DataVol1" (with virtual size 5,42 TiB)
/dev/vg1/lv2 - it is QNAP "DataVol2" (with virtual size 512 GiB)

So my data are on /dev/vg1/lv1 and /dev/vg1/lv2 but they are "inactive", 
so they cannot be mounted.
$ sudo mount /dev/vg1/lv2 /mnt/lv2
mount: special device /dev/vg1/lv2 does not exist

I tried to activate it with lvchange but there is message that manual 
repair of vg1/tp1 is required:
$ sudo lvchange -ay vg1/lv2
   Check of pool vg1/tp1 failed (status:1). Manual repair required!

I tried command lvconvert --repair vg1/tp1 with no success:
$ sudo lvconvert --repair vg1/tp1
   Using default stripesize 64,00 KiB.
   Volume group "vg1" has insufficient free space (0 extents): 4096 
required.

Please help howto repair/mount this LVM...

Best Regards,
Daniel

Re: [linux-lvm] Problem with mount QNAP disks in Linux

[Danniello]

W dniu 12.06.2017 o 15:42, Marian Csontos pisze:
>> $ sudo vgs
>>    VG     #PV #LV #SN Attr   VSize   VFree
>>    vg1      1   4   0 wz--n-   5,44t    0
>
> VFree 0 - that's bad...
>
>>
>> I tried to activate it with lvchange but there is message that manual 
>> repair of vg1/tp1 is required:
>> $ sudo lvchange -ay vg1/lv2
>>    Check of pool vg1/tp1 failed (status:1). Manual repair required!
>>
>> I tried command lvconvert --repair vg1/tp1 with no success:
>> $ sudo lvconvert --repair vg1/tp1
>>    Using default stripesize 64,00 KiB.
>>    Volume group "vg1" has insufficient free space (0 extents): 4096 
>> required.
>
> Repair needs some space to write new metadata, and the message says it 
> all: there is no free space in the volume group.
>
> Add more space to the volume group: vgextend vg1 DEVICE. You will 
> either need to add a disk, or carve out some space out of other MD 
> devices.
>
> Or free some space in the VG: you can not shrink thin pool, so the 
> only other option is, if the data on vg1/lv544 are not interesting or 
> can be moved elsewhere, you could remove that and let repair use that 
> space.
>
> Also I noticed the size of logical volume lv1 alone is 5,42t (and with 
> lv2 it is approximately 5,9t) while thin pool's is only 5,40t: that is 
> not a good setup - it will eventually overfill (maybe it already did) 
> and you will not be able to resize the pool any further to accommodate 
> all data - with older kernels this was a serious problem and could 
> lead to a file system corruption.
>
> If that's what the NAS created for you, it should be reported to the 
> manufacturer as well.
>
> Once you can mount the volumes, I strongly recommend shrinking lv1. 

This setup was done by me on QNAP QTS - at beginning I had only 
DataVol1, but after some time I added "small" DataVol2. It is not good 
configuration, but in QTS system there was no option to shrink DataVol1.
Anyway it was working OK in QNAP QTS system (but with warnings that pool 
is almost full). Unfortunately my QNAP is no longer working - 
motherboard defected...

I have backup of all important data on other places, but wanted to try 
restore also "not important data":)

In QNAP QTS system it should work without additional actions - only 
after unlock encrypted DataVol1 and DataVol2 there should be message 
about check filesystem. I do not have access to other working QNAP, so I 
tried with my desktop system with Fedora, but activate lv1 and lv2 was 
not working on default configuration.

Probably you are right - I should add storage and then try to repair it. 
But I found "workaround" by modifying /etc/lvm/lvm.conf in global section:
thin_check_executable = ""

I know that disable thin_check is generally not recommended, but I 
wanted try everything before do more drastic methods. After this change 
I could activate volumes lv1 and lv2:
lvchange -ay /dev/vg1

Decrypt:
cryptsetup luksOpen /dev/vg1/lv1 crypt_lv1
cryptsetup luksOpen /dev/vg1/lv2 crypt_lv2

And mount:
mount -r /dev/mapper/crypt_lv1 /mnt/lv1
mount -r /dev/mapper/crypt_lv2 /mnt/lv2
Data accessible:)

I want only copy some data from it, so repair will not be necessary. 
Next I plan build my own NAS, so I will reformat disks from scratch.

Thank you!
Daniel