Make sure the filesystem isn’t being used:
sudo systemctl stop plexmediaserver
sudo systemctl stop smbd
unmount the filesystem
sudo umount /storage
use parted to resize the filesystem
sudo parted
resize part 1 100%
remount
sudo mount /storage
resize the Btrfs filesystem
sudo btrfs filesystem resize max /storage
and, all being well, the system isn’t hosed.
Quick overview more for me than anything else:
lsblk to see disks in system:
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sda 8:0 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdb 8:16 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdc 8:32 0 3.6T 0 disk
├─sdc1 8:33 0 1G 0 part
├─sdc2 8:34 0 1G 0 part
│ └─md0 9:0 0 1022M 0 raid1
│ └─md0p1 259:2 0 1020M 0 part /boot
├─sdc3 8:35 0 128G 0 part
│ └─md1 9:1 0 127.9G 0 raid1
│ └─md1p1 259:1 0 127.9G 0 part [SWAP]
└─sdc4 8:36 0 3.5T 0 part
└─md2 9:2 0 3.5T 0 raid1
└─md2p1 259:3 0 3.5T 0 part /
sdd 8:48 0 3.6T 0 disk
├─sdd1 8:49 0 1G 0 part /boot/efi
├─sdd2 8:50 0 1G 0 part
│ └─md0 9:0 0 1022M 0 raid1
│ └─md0p1 259:2 0 1020M 0 part /boot
├─sdd3 8:51 0 128G 0 part
│ └─md1 9:1 0 127.9G 0 raid1
│ └─md1p1 259:1 0 127.9G 0 part [SWAP]
└─sdd4 8:52 0 3.5T 0 part
└─md2 9:2 0 3.5T 0 raid1
└─md2p1 259:3 0 3.5T 0 part /
sde 8:64 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdf 8:80 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
add in new disk and lsblk again:
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sda 8:0 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdb 8:16 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdc 8:32 0 3.6T 0 disk
├─sdc1 8:33 0 1G 0 part
├─sdc2 8:34 0 1G 0 part
│ └─md0 9:0 0 1022M 0 raid1
│ └─md0p1 259:2 0 1020M 0 part /boot
├─sdc3 8:35 0 128G 0 part
│ └─md1 9:1 0 127.9G 0 raid1
│ └─md1p1 259:1 0 127.9G 0 part [SWAP]
└─sdc4 8:36 0 3.5T 0 part
└─md2 9:2 0 3.5T 0 raid1
└─md2p1 259:3 0 3.5T 0 part /
sdd 8:48 0 3.6T 0 disk
├─sdd1 8:49 0 1G 0 part /boot/efi
├─sdd2 8:50 0 1G 0 part
│ └─md0 9:0 0 1022M 0 raid1
│ └─md0p1 259:2 0 1020M 0 part /boot
├─sdd3 8:51 0 128G 0 part
│ └─md1 9:1 0 127.9G 0 raid1
│ └─md1p1 259:1 0 127.9G 0 part [SWAP]
└─sdd4 8:52 0 3.5T 0 part
└─md2 9:2 0 3.5T 0 raid1
└─md2p1 259:3 0 3.5T 0 part /
sde 8:64 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdf 8:80 0 9.1T 0 disk
└─md3 9:3 0 27.3T 0 raid5
└─md3p1 259:0 0 27.3T 0 part /storage
sdg 8:96 0 9.1T 0 disk
├─sdg1 8:97 0 2G 0 part
│ └─md127 9:127 0 0B 0 md
├─sdg2 8:98 0 9.1T 0 part
│ └─md126 9:126 0 0B 0 md
├─sdg3 8:99 0 1G 0 part
└─sdg4 8:100 0 1G 0 part
/dev/sdg has been added
However, because sdg is showing as md127 we need to stop and remove the raid:
mdadm --stop /dev/md127
mdadm --remove /dev/md127
now we can partition the disk
use parted to wipe filesystem and add a new label to the disk
parted -s -a optimal /dev/sdg mklabel gpt
Use fdisk to check the new GTP Partition tables. Don’t bother to create any partitions.
fdisk -p to check out the unpartitioned disk:
$ sudo fdisk /dev/sdg
Welcome to fdisk (util-linux 2.39.3).
Changes will remain in memory only, until you decide to write them.
Be careful before using the write command.
Command (m for help): p
Disk /dev/sdg: 9.1 TiB, 10000831348736 bytes, 19532873728 sectors
Disk model: WDC WD100EFAX-68
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: gpt
Disk identifier: 56FF5A1F-FE90-4AB1-ABD6-DBB1E1D22042
Command (m for help): q
Just check to see which array you’re adding the disk to:
$ cat /prod/mdstat
Personalities : [raid6] [raid5] [raid4] [raid1] [raid0] [raid10]
md1 : active raid1 sdd3[1] sde3[0]
134085632 blocks super 1.2 [2/2] [UU]
bitmap: 0/1 pages [0KB], 65536KB chunk
md2 : active raid1 sdd4[1] sde4[0]
3770516480 blocks super 1.2 [2/2] [UU]
bitmap: 1/29 pages [4KB], 65536KB chunk
md0 : active raid1 sdd2[1] sde2[0]
1046528 blocks super 1.2 [2/2] [UU]
md3 : active raid5 sdf[2] sdb[1] sdg[4] sda[0]
29298914304 blocks super 1.2 level 5, 512k chunk, algorithm 2 [4/4] [UUUU]
bitmap: 0/73 pages [0KB], 65536KB chunk
unused devices: <none>
md3 in this case. Add the disk into the array:
mdadm --add /dev/md3 /dev/sdg
Check the array:
$ sudo mdadm --add /dev/md3 /dev/sdg
mdadm: added /dev/sdg
sysadmin@tnas:~$ cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid1] [raid0] [raid10]
md1 : active raid1 sdd3[1] sde3[0]
134085632 blocks super 1.2 [2/2] [UU]
bitmap: 0/1 pages [0KB], 65536KB chunk
md2 : active raid1 sdd4[1] sde4[0]
3770516480 blocks super 1.2 [2/2] [UU]
bitmap: 2/29 pages [8KB], 65536KB chunk
md0 : active raid1 sdd2[1] sde2[0]
1046528 blocks super 1.2 [2/2] [UU]
md3 : active raid5 sdc5 sdf[2] sdb[1] sdg[4] sda[0]
29298914304 blocks super 1.2 level 5, 512k chunk, algorithm 2 [4/4] [UUUU]
bitmap: 0/73 pages [0KB], 65536KB chunk
The disk is now visible in the array: sdc5 but it’s marked as a spare disk (S).
Now we need to grow the array:
mdadm --grow /dev/md3 --raid-devices=5
where the number of raid devices is the number of disks we want in the raid.
Watch /proc/mdstat to see the progress:
$ cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid1] [raid0] [raid10]
md1 : active raid1 sdd3[1] sde3[0]
134085632 blocks super 1.2 [2/2] [UU]
bitmap: 0/1 pages [0KB], 65536KB chunk
md2 : active raid1 sdd4[1] sde4[0]
3770516480 blocks super 1.2 [2/2] [UU]
bitmap: 1/29 pages [4KB], 65536KB chunk
md0 : active raid1 sdd2[1] sde2[0]
1046528 blocks super 1.2 [2/2] [UU]
md3 : active raid5 sdc[5] sdf[2] sdb[1] sdg[4] sda[0]
29298914304 blocks super 1.2 level 5, 512k chunk, algorithm 2 [5/5] [UUUUU]
[>………………..] reshape = 0.0% (337408/9766304768) finish=3376.6min speed=48201K/sec
bitmap: 0/73 pages [0KB], 65536KB chunk
unused devices: <none>
Yep, this is going to take a while. Time to put the kettle on.
Helpful guides:
https://wpguru.co.uk/2021/01/expand-software-raid-mdadm/
https://www.itsfullofstars.de/2019/03/how-to-add-a-new-disk-to-raid5/
https://tomlankhorst.nl/setup-lvm-raid-array-mdadm-linux
Edit /etc/netplan/50-cloud-init.yaml
network:
ethernets:
enp3s0:
dhcp4: true
dhcp6: true
mtu: 1472
version: 2
Then do:
sudo netplan try
and accept the change if it seems to work.
It appears that the coreutils developers have gone rogue and changed the default format of the ‘ls’ command, instead of being sensible and showing a filename like this:
What's New Scooby-Doo? (2002)
it now shows this:
'What'\''s New Scooby-Doo (2002)'
The fix is to add:
export QUOTING_STYLE=literal
to the .bashrc file.
My reasons for putting Ubuntu are explained here: Confession Time and I’m using Ubuntu because that’s what Terra-Master use themselves so it should work just fine and it’s the distribution I’m slightly more familiar with at the moment. Debian would be my alternative choice.
First things first. Open up the case and remove the small SanDisk USB stick. This contains the TOS boot loader and we don’t want that anymore. Go and put it in a marked envelope in a drawer somewhere just in case you ever want to put it back.
Step 2 – not for the faint of heart. Update the BIOS. The full thread is here: https://forum.terra-master.com/en/viewtopic.php?t=4238 and with that done the rest is pretty straightforward.
Next we need to change the boot order. With the TOS USB stick removed that’s an awful lot easier. For some reason, for me at least, the TOS USB stick always took priority no matter what I did so with that out of the way it’s a lot easier to do things now.
I actually burned Ubuntu to DVD rather than a flash drive simply because I don’t think the flash drives I have are very reliable. They are incredibly slow for some reason too. So, I opted for DVD.
Slap the DVD in and run through the basic setup until the drive configuration page presents itself. Here choose the custom layout option. I have two Seagate IronWolf disks as my boot disks so I wanted a mirrored raid boot setup. Not having done that before I found a very easy to follow video (and I really hate having to follow videos) here:
Installing Ubuntu 24.04 With RAID
and a really useful page here:
Ubuntu 24.04 with software RAID1 and UEFI
The only changes to the process I made was to include in my four 10TB raid disks into the process. That means that at the end of the video I have a filesystem that looks something like:
md0 RAID1 /boot
md1 RAID1 SWAP
md2 RAID1 /root
md3 RAID5 /storage
It also means that as the storage is built at installation time I can just sit back and let the RAID configure itself for a few hours.
Because I always choose a minimised installation I need to add a few things in to make the environment usable so this is what I do:
sudo apt update
sudo apt install htop
sudo apt install avahi-daemon
sudo apt install vim
sudo apt install cron
sudo apt upgrade
Once cron and vim are installed I set up my lazy mdstat rebuild page.
I have a little web server with a basic html page of:
<!DOCTYPE html>
<html>
<head>
<title>
RAID Rebuild Status
</title>
<meta http-equiv="refresh" content="60" >
</head>
<body>
<embed src="raid.txt" width="100%" height="1000">
</body>
</html>
It’s really not much of a page. Then on the T12 I add in a cron job:
* * * * * cat /proc/mdstat | ssh ricardo@server.local 'cat - > /home/ricardo/docker/dumkid/html/raid.txt'
This means that I can point my phone / laptop / work computer to my little webpage and see how the rebuild is going without having to log in to the T12 itself.
Terra-Master seems to use mdadm:
TNAS:~# ps -ef | grep [m]dadm
davros 547 1 0 Jan12 ? 00:00:00 /sbin/mdadm --monitor --scan --daemonise --syslog --delay=10 --program=/etc/init.d/nas/mdadm-monitor
so, it looks like the best way to follow the rebuild is:
TNAS:~# watch cat /proc/mdstat
which will generate an output like:
Every 2.0s: cat /proc/mdstat TNAS: Mon Jan 13 09:19:44 2025
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath] [faulty]
md1 : active raid1 sdd4[1] sda4[0]
9755199616 blocks super 1.2 [2/1] [U_]
[===>.................] recovery = 18.9% (1852244544/9755199616) finish=667.2min speed=197414K/sec
md0 : active raid1 sdc4[1] sdb4[0]
3895824636 blocks super 1.2 [2/2] [UU]
md8 : active raid1 sdc3[1] sdb3[0]
1997824 blocks super 1.2 [2/2] [UU]
bitmap: 0/1 pages [0KB], 65536KB chunk
md9 : active raid1 sdc2[2] sdb2[0]
7995392 blocks super 1.2 [2/2] [UU]
bitmap: 0/1 pages [0KB], 65536KB chunk
unused devices: <none>
This was a new thing I came across this morning:
2 devices have a firmware upgrade available.
Run `fwupdmgr get-upgrades` for more information.
I’ve not seen that before so I had to figure out how to fix the issue. Turns out it’s pretty simple:
fwupdmgr get-upgrades
Shows the available upgrades
fwupdmgr update
Applies any updates.
As much as I love Docker (even though I’m not brilliant with it) sometimes I just need a proper server. In this case I need a nice shiny WebDav server with the ability to show web pages too.
I’m using Ubuntu for this because, as much as I like Debian, I find Ubuntu just that little bit easier to work with. Except Snap. Snap can burn in hell. The rest of it is pretty nice though.
The plan is to use Nginx as the web /WebDav server and if I can configure it up the way I need and get it doing everything I want I can then get the UNIX team at work to build me one for the office. I can’t help but wonder if they can manage that.
I can never remember how to do this and the Ubuntu pages are, surprisingly, useless. So, for future reference:
In my case I’m using a small Dell OptiPlex 7070 Micro. So, to get into the BIOS it’s:
It should now be possible to boot from the USB stick.
When partitioning the disk don’t get confused between the USB stick filesystem and the PC filesystem. It’s pretty easy once you realise that the PC is NVME.
