MDADM didn't grow after adding disk on Debian 8

[Melkor]

Hello there,

I'm encountering an issue with my personal custom NAS on Debian 8, and I hope someone here will be geek enough to have some clue...
The NAS has a RAID5 mounted on using mdadm (no spare drive). I had a 6x 3To disks array, but I was too tight, so I added another 3To disk.
I did as usual (I started this RAID with 4 HDD and already added 2 more HDD) and typed theses commands:

sgdisk -R=/dev/sdi /dev/sdb
sgdisk -G /dev/sdi

mdadm --manage /dev/md0 --add /dev/sdi1
mdadm --grow --raid-devices=7 --backup-file=/root/grow_md0.bkp /dev/md0

xfs_growfs /dev/md0

Everything did go as usual, and after about 30 hours, the last job (the long one) did finish.

But now when I'm checking the array volume, I'm still with 14To I had before instead of the 17ish I should get.

After some thinking, the only thing I did differently this time, was a wrong command:

mdadm --grow --raid-devices=6 --backup-file=/root/grow_md0.bkp /dev/md0

instead of

mdadm --grow --raid-devices=7 --backup-file=/root/grow_md0.bkp /dev/md0

but I got an error message and then typed the right one.

Here are some logs:

# cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4]
md0 : active raid5 sdb1[5] sdg1[9] sdh1[8] sdi1[7] sde1[4] sdd1[2] sdc1[6]
17581581312 blocks super 1.2 level 5, 512k chunk, algorithm 2 [7/7] [UUUUUUU]

unused devices: <none>

# mdadm -D /dev/md0
/dev/md0:
Version : 1.2
Creation Time : Fri Oct 19 03:15:33 2012
Raid Level : raid5
Array Size : 17581581312 (16767.10 GiB 18003.54 GB)
Used Dev Size : 2930263552 (2794.52 GiB 3000.59 GB)
Raid Devices : 7
Total Devices : 7
Persistence : Superblock is persistent

Update Time : Fri Jan 3 15:41:29 2020
State : clean
Active Devices : 7
Working Devices : 7
Failed Devices : 0
Spare Devices : 0

Layout : left-symmetric
Chunk Size : 512K

Name : Melkor-Server:0
UUID : c1335b52:217970e8:407df77c:199a66ca
Events : 38463

Number Major Minor RaidDevice State
5 8 17 0 active sync /dev/sdb1
6 8 33 1 active sync /dev/sdc1
2 8 49 2 active sync /dev/sdd1
4 8 65 3 active sync /dev/sde1
7 8 129 4 active sync /dev/sdi1
8 8 113 5 active sync /dev/sdh1
9 8 97 6 active sync /dev/sdg1

It would be a great new year gift for me to have my array regain it right size, so thank by advance!


EDIT:

It seems that the array is in fact the right size, but not the partition (even after xfs_growfs). Here are some more logs:

# parted -l

Model: Linux Software RAID Array (md)
Disk /dev/md0: 18.0TB
Sector size (logical/physical): 512B/4096B
Partition Table: loop
Disk Flags:

Number Start End Size File system Flags
1 0.00B 18.0TB 18.0TB xfs

[...]

# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/md0 14T 14T 110G 100% /media/NAS
[...]

[p.H]

17581581312 blocks super 1.2 level 5, 512k chunk, algorithm 2 [7/7] [UUUUUUU]
Array Size : 17581581312 (16767.10 GiB 18003.54 GB)

Where do you see 14 TB ?

[Melkor]

Hi,

Thanks for the answer.

Since my array didn't grow after adding my last HDD, I supposed this meant the total volume before subtracting the Dev Size (the 3To of parity data).
Unfortunately, I don't recall what was showing this command before adding the disk :/

If the command output is right, how is it possible for me to not have more usable space after the process?

Thanks again for your time.

[Bloom]

After growing the RAID array, you must also grow the filesystem present on that array. That doesn't happen automatically, you must do that manually.

[Melkor]

Isn't that what this command do?

xfs_growfs /dev/md0

[Melkor]

17581581312 blocks super 1.2 level 5, 512k chunk, algorithm 2 [7/7] [UUUUUUU]
Array Size : 17581581312 (16767.10 GiB 18003.54 GB)

Where do you see 14 TB ?

I added more commands output in my original post, like df -h which shows 14To.

Is it possible that the xfs_growfs command didn't work? And in that case is it useful to rerun it?

[p.H]

Is it possible that the xfs_growfs command didn't work?

Maybe. I have no experience with XFS, but the man page says that it expects the mount point, not the device file, and -d to use the maximum size.

[Melkor]

Hi again,

I did execute xfs_growfs -d /dev/md0 and it did work.

Thank you for your time and leads, and have a nice year!