1 TB SSD NVME - only 800 GiB used

No20 · July 4, 2021, 1:19am

Hi,

I would like to know if I should file a bug report.
I have a new notebook with a 1 TB NVME SSD. I installed Qubes and restored the backup from my old notebook. My old notebook has a 1 TB SSD Micro Sata. On there 7,7 GiB are swap and 907 GiB is for the dom0-pool logical volume.
On my new notebook the dom0 logical volume only uses about 800 GiB. The swap is 30 GiB as can bee seen from the output of lsblk:

lsblk
NAME                                       MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
nvme0n1                                    259:0    0 931.5G  0 disk  
├─nvme0n1p1                                259:1    0     1G  0 part  /boot
└─nvme0n1p2                                259:2    0 930.5G  0 part  
  └─luks-e27da347-bd2e-4e9a-b1d4-f48d145ee9b9
    253:0    0 930.5G  0 crypt 
    ├─qubes_dom0-pool00_tmeta              253:1    0   816M  0 lvm   
    │ └─qubes_dom0-pool00-tpool            253:3    0 798.1G  0 lvm
....
    └─qubes_dom0-swap                      253:5    0  30.8G  0 lvm   [SWAP]
loop0                                        7:0    0     1G  0 loop

I know that there is a conversion reduction from GB to GiB.
Can somone explain to me why Qubes is not using the full space? Could it have something to do with restoring the backup? Should I file a bug report?

augsch · July 4, 2021, 1:49am

github.com/QubesOS/qubes-issues

Enable LVM autoextend

opened 10:19PM - 15 May 20 UTC

closed 03:12AM - 26 Jun 21 UTC

marmarek

C: installer P: default T: bug

**Qubes OS version** R4.0 **Affected component(s) or functionality** lvm, i…nstaller **Brief summary** Installer (anaconda) leaves some space not allocated (~20% now, previously less). This is supposed to be used by LVM autoextend to automatically enlarge data or metadata volumes according to the needs. But autoextend is disabled. **To Reproduce** _Steps to reproduce the behavior:_ _1. Install Qubes 4.0_ _2. Check LVM pool size and settings_ _3. See ~20% not allocated, but `thin_pool_autoextend_threshold` in `/etc/lvm/lvm.conf` set to 100 (effectively disabling the feature)_ **Expected behavior** When free space is left, autoextend should be enabled to make use of it. **Actual behavior** Autoextend not enabled. **Additional context** This is yet another countermeasure for filling up metadata volume. **Relevant [documentation](https://www.qubes-os.org/doc/) you've consulted** #3243 **Related, [non-duplicate](https://www.qubes-os.org/doc/reporting-bugs/#new-issues-should-not-be-duplicates-of-existing-issues) issues** _A list of links to other bug reports, feature requests, or tasks in the qubes-issues tracker (or "none" if you didn't find any). Do not describe any other unreported bugs, features, or tasks here._

This issue may be related to your problem.

No20 · July 4, 2021, 2:41am

Thanks augsch. It definitely sounds like it. Any possibility to extend the size manually?

fepitre · July 4, 2021, 9:29pm

Yes check what we do in %post: Enable LVM autoextend and set default value to 90 for thin_pool · fepitre/qubes-core-admin-linux@605cb47 · GitHub

No20 · July 6, 2021, 12:27am

So I would need to do a full reinstall when 4.1 is released?

51lieal · July 6, 2021, 1:35am

You need to know how much space you have.
Extend the vm–pool

example :

[root@dom0 51lieal]# vgs
  VG  #PV  #LV  #sn Attr  VSize  VFree
  qubes_dom0  2  60  0  wz--n- 1.37t  99.97g

[root@dom0 51lieal]# lvextend -L +99.97GiB /dev/mapper/qubes_dom0-vm--pool -r

Wait for some time, and then your vm-pool wlll extend.

No20 · July 6, 2021, 10:31am

Thanks for your reply. I am not comfortable with using lvextend. I once did break my install with it. Right now I don’t need the extra space but thanks anyway

51lieal · July 6, 2021, 10:40am

I have tested it, just take a note about your vg and data-pool.
Just to make sure, if you can provide full lsblk i can give you a hint.
And you can always train qubes in vm.