- Using the qubes 4.1-beta installer, it is very difficult to add a standard ext4 partition to the end of the default LVM setup. Choosing Custom, if you try to resize down the existing LVM volumes it complains about requiring a mount point. Afterwards, you can’t add a standard partition anyway. I could just be an LVM novice.
- In the
lvs
output of the installed system, vm-pool (676g) + root-pool (20g) + swap (4g) doesn’t add up to the size of /dev/sda2 (800g). Is it normal to have 100g unaccounted for? This is the result of pre-making a standard partition at the end of the disk and letting the qubes installer automatically partition the remaining free space.
I’m not sure what you mean, if you could explain a bit what are you trying to achieve.
try read this 1 TB SSD NVME - only 800 GiB used - User Support - Qubes OS Forum (qubes-os.org)
My goal was to add an extra standard partition at the end of the usual qubes partitions:
/dev/sda 1TB SSD
/dev/sda1 boot
/dev/sda2 encrypted volume containing LVM pools
/dev/sda3 128 GiB ext4
Creating /dev/sda3 above was really hard / impossible to do using installer’s Custom partitioning option (using the 4.1-beta iso). (Maybe it could be done with the Advanced option but I didn’t know how things are laid out in 4.1, scared me off).
try read this 1 TB SSD NVME - only 800 GiB used - User Support - Qubes OS Forum (qubes-os.org)
If I understand correctly, /dev/sda2 is deliberately under-utilized, and the vm-pool and root-pool will automatically grow into the unused space if one gets more than 90% full? I guess this solves the problem of one of vm-pool, root-pool, or their respective metadata volumes running out of space before the others, as long as there is space left in /dev/sda2?
-
After reach installer ctrl + alt + f2
-
Wipe disk and create partition using gdisk / fdisk (boot, lvm, and 128 GiB ext4)
-
boot (sda1) lvm (sda2) 128gib (sda3)
-
cryptsetup -c aes-xts-plain64 -h sha512 -s 512 --use-random -y -i 10000 luksFormat /dev/sda2
-
crypsetup luksOpen /dev/sda2 luks
-
pvcreate /dev/mapper/luks
-
vgcreate qubes_dom0 /dev/mapper/luks
-
lvcreate -n swap -L 4G qubes_dom0
-
lvcreate -T -L 20G qubes_dom0/root-pool
-
lvcreate -T -l +100%FREE qubes_dom0/vm-pool
-
lvs
-
lvcreate -V20G -T qubes_dom0/root-pool -n root
-
lvcreate -V( size in your vm pool )G -T qubes_dom0/vm-pool -n vm
-
mkfs.ext4 /dev/qubes_dom0/vm
back to installer gui with ctrl + alt + f6
setup everthing and in storage, refresh disk first and hit rescan, use custom click done.
reformat boot, root, swap, and leave qubes_dom0/vm.
proceed with installation, and you’re done with another 128gb untouched.
check my featured topic for detailed instructions.
If I understand correctly, /dev/sda2 is deliberately under-utilized, and the vm-pool and root-pool will automatically grow into the unused space if one gets more than 90% full? I guess this solves the problem of one of vm-pool, root-pool, or their respective metadata volumes running out of space before the others, as long as there is space left in /dev/sda2?
yes