Btrfs lsblk messed up after running btrfs-assistant (with snapper installed)

[jackdinn]

I’m still tinkering with btrfs

I am looking at snapper again. Im on a clean OS installation of KDE on a VBox.
I only installed snapper and btrfs assistant.

I need to know what is going on here?

I show my blocks:

>lsblk                                                                                                                                                      
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
sda      8:0    0   40G  0 disk 
├─sda1   8:1    0   40G  0 part /var/log
│                               /var/cache
│                               /home
│                               /
└─sda2   8:2    0    8M  0 part 
sr0     11:0    1 1024M  0 rom  

I run btrfs-assistant, click around a few tabs

>lsblk                                                                                  
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
sda      8:0    0   40G  0 disk 
├─sda1   8:1    0   40G  0 part /run/BtrfsAssistant/c863c9b5-d87a-4196-bdfb-5002d20e4ce0
└─sda2   8:2    0    8M  0 part 
sr0     11:0    1 1024M  0 rom  

Then close btrfs-assistant. I have done nothing in it of any consequence. Just opened it, looked at the tabs and closed it.

Where on earth are my btrfs mounts? it says even /@ is not mounted but of course it must be along with all the others?

>lsblk                                                                                                                                                       
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
sda      8:0    0   40G  0 disk 
├─sda1   8:1    0   40G  0 part /var/log
└─sda2   8:2    0    8M  0 part 
sr0     11:0    1 1024M  0 rom  

Mounts look ok

>mount | grep -i btrfs                                                                  
/dev/sda1 on / type btrfs (rw,relatime,space_cache=v2,subvolid=256,subvol=/@)
/dev/sda1 on /home type btrfs (rw,relatime,space_cache=v2,subvolid=257,subvol=/@home)
/dev/sda1 on /var/cache type btrfs (rw,relatime,space_cache=v2,subvolid=258,subvol=/@cache)
/dev/sda1 on /var/log type btrfs (rw,relatime,space_cache=v2,subvolid=259,subvol=/@log)

Whats going on? and how do i get the lsblk mount list back without a reboot?

P.S. If snapper is not installed this does not happen.

[andreas85]

@jackdinn

mount | grep -i btrfs

You can also use:

mount -t btrfs

@jackdinn

Where on earth are my btrfs mounts? it says even /@ is not mounted but of course it must be along with all the others?

I think this has nothing to do with btrfs-mounts, but only with the lsblk command itself

[jackdinn]

@andreas85

I think this has nothing to do with btrfs-mounts, but only with the lsblk command itself

So what are you saying? I can just ignore lsblk? But when I’m messing about with btrfs (and generally messing with file systems) I need to use lsblk.

[Nachlese]

@jackdinn

and how do i get the lsblk mount list back without a reboot?

in similar situations, when using gparted to change stuff around, for instance
partprobe
is used to inform the kernel of any possible changes

lsblk -f would perhaps be more useful than just plain lsblk

@jackdinn

I need to use lsblk.

Do you?
How are you so sure?

you know you are dealing with btrfs
btrfs has, from what I have seen here, it’s own dedicated set of tools

[jackdinn]

@Nachlese

lsblk -f would perhaps be more useful than just plain lsblk

yea lsblk -f and lsblk -m is the kind of lsblk command i would be using but they are also messed up.

@Nachlese

you know you are dealing with btrfs
btrfs has, from what I have seen here, it’s own dedicated set of tools

Yes it does, but i still use lsblk for quick UUID info and to check mounts quick.

Look, I just want to know how to get lsblk to “refresh” itself, i dunno what it would be called but it needs to look at the mounts fresh. How do i do that?

[Nachlese]

No idea.
I stay away from btrfs because it is so different than anything else.

@jackdinn

i dunno what it would be called but it needs to look at the mounts fresh.

and I don’t even know what you might mean by that

@jackdinn

but i still use lsblk for quick UUID info and to check mounts quick.

dito - it might just be the wrong tool for the job

I just don’t know.
(I mentioned partprobe …)

[Kobold]

@Nachlese

No idea.
I stay away from btrfs because it is so different than anything else.

+1
btrfs is just next level to run into a bugs festival, i mean if he like to fight bugs… why not?

[jackdinn]

LOL, yeah maybe you are right, but I was of the opinion that it was to help me save the headaches & system mess-ups in snapshot states that will allow me to roll back the system when I need to.

(I had just one too many system updates that borked my system. I’m still recovering from the Plasma 6 “stable” update).

After managing to get most things working again, I decided I would take a look at snapshot file systems. I would have rather used ZFS as I use that on my NAS, but it seemed that BTRFS was more “implemented” (has more apps and is built into the installer etc).

I’ve been at it for a few weeks now, we shall see

[Kobold]

@jackdinn

I was of the opinion that it was to help me save the headaches & system mess-ups in snapshot states that will allow me to roll back the system when I need to.

(I had just one too many system updates that borked my system. I’m still recovering from the Plasma 6 “stable” update).

I just used Timeshift Rsync Snapshots on my PC+Laptop’s both with ext4 installed and never had any problems from a rollback when using the GUI from a Live Distro… it was childish easy to restore it.

What application did you used for creating snapshots? Maybe you did a simple user failure?

[jackdinn]

@Kobold

What application did you used for creating snapshots? Maybe you did a simple user failure?

I never had a problem with any snapshots (yet), not sure what you mean.

[Kobold]

I don’t understand, i thought you complained in my quote above, that you using btrfs now to evade a issue that you had with rolling your system back?

[jackdinn]

No, im now looking at snapshot file systems to evade system updates causing destruction of my computer. (and user errors as well).

You said that btrfs is the cause of many headaches and i was saying that i thought these snapshot type filesystems were exactly to prevent these kind of system mess-ups from being a headache.

I have had one to many system updates that had borked my system, so i decided to start looking into this type of snapshot filesystem.

It’s been working fine in testing and im now using btrfs as my main systems FS.

Im still pushing the testing hard, and looking at all the apps/packages that can be used (although a lot seem to be out of date or of no use nowadays).

I already use rsync, borg, vorta, dd, partition images etc etc, i just cant help from tinkering with stuff i probably shouldn’t.

Like i say, we shall see how it goes.

[Kobold]

When i see recently that you can meld 2 btrfs partition together and that you can do overlapping mistakes and stuff like that… it let me worrie how userfriendly btrfs really is.

I would probably testing it year’s before i use it on my main system… but even then i think my files are to important to me and i probably would evade it.

At the end of the day, all Filesystems has their strength and weakness.

[jackdinn]

@Kobold

I would probably testing it year’s before i use it on my main system… but even then i think my files are to important to me and i probably would evade it.

Yea i hear you, and sure my home folder is 100% important. However, i do have so many backups on my NAS, on my remote server, on disconnected backup drives. I do not see why you would not jump in and use things that you are wanting to learn (well not immediately but after you are somewhat comfortable with them). I mean all my data is backup in multiple ways. The only way to learn is to use whatever you are leaning on a daily basis, otherwise it just does not sink in (well does not in my case, im not a spring chicken any longer).
Use it or loose it ^^

If i like btrfs over other standard incremental file transfer type backups then i will just keep using it, if not i will drop back to just borg (my main NAS backup method).

[Kobold]

Here check this Topic out, it has some good info about btrfs and snapshots already:

BTRFS with ‘/home’ on separate disk

I have a Manjaro (v23) installation with the following configuration: / mounted on my SSD, which is /dev/sdb3 /home mounted on my HDD, which is /dev/sdc1 This way: sda 8:0 0 931.5G 0 disk └─sda1 8:1 0 931.5G 0 part sdb 8:16 0 223.6G 0 disk ├─sdb1 8:17 0 1G 0 part /boot/efi ├─sdb2 8:18 0 6G 0 part [SWAP] └─sdb3 8:19 0 216.6G 0 part /var/log /var/cache /swap / sdc 8:32 0 931.5G 0 disk └─sdc1 8…

[jackdinn]

Yea, but a bit out of date (i didn’t read it all)? The /@home is created and mounted automatically by the installer. Timeshift gives you a simple option to snapshot your home folder if you wish and more importantly give you the option to restore or not to restore the home snapshot when doing system restore. I did not have to worry about any of what they are talking about, not with timeshift anyhow, its was all working as one might expect, “out of the box”

I wrote a script and background services that send btrfs snapshots created by timeshift to an external device incrementally. This has been working for a week now. It sends both / and /home snapshots over. I have tested bringing them back and using them to restore a system successfully ^^ (yea i could do it with just files but the speed is night and day).

Im looking into using snapper and removing timeshift though (because reasons), but i don’t foresee to many problems.

I think the post you linked had problems because he had his home mounted from a separate device, I think using snapper would have made his life easier.

Like i say, i just cant help myself but to tinker

P.S. Or even dont use any of these apps, just use the btrfs tools and create your own snapshot system. I might go down that road after looking at all the apps already created.

[Molski]

@jackdinn

LOL, yeah maybe you are right, but I was of the opinion that it was to help me save the headaches & system mess-ups in snapshot states that will allow me to roll back the system when I need to.

I wouldn’t let the opinion of one dissuade you. To get to the point where you could boot a live environment, and type the commands manually to rollback; is a little extra reading, but it’s not terrible. The information is very abundant.

@jackdinn

(I had just one too many system updates that borked my system. I’m still recovering from the Plasma 6 “stable” update).

If you’re following the stable release posts, and dealing with pacnew files properly and whatnot; this should be a pretty rare occurrence.

@jackdinn

After managing to get most things working again, I decided I would take a look at snapshot file systems. I would have rather used ZFS as I use that on my NAS, but it seemed that BTRFS was more “implemented” (has more apps and is built into the installer etc).

I’ve used zfs a lot. For better performing and generally larger arrays, it’s still my choice. But there’s a lot of cool stuff you can do with btrfs that zfs can’t.

For backup purposes and the block level incremental sending of snapshots, it is really handy here if your “home folder is 100%” important. Once setup properly, it is way more efficient, and a lot more reliable versus something like rsync.

Why your lsblk looks like that is weird. And you don’t even get UUIDs with -f? Also what kernel are you running?

[jackdinn]

The kernel is 6.9.2-1-MANJARO

here is lsblk, before i open btrfs-assistant and then after, tested with all (well not all) switches.

    ~  lsblk                                                                                                                                                       ✔ 
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
sda      8:0    0   40G  0 disk 
├─sda1   8:1    0   40G  0 part /var/log
│                               /var/cache
│                               /home
│                               /
└─sda2   8:2    0    8M  0 part 
sr0     11:0    1 1024M  0 rom  
    ~  uname                                                                                                                                                       ✔ 
Linux
    ~  uname -a                                                                                                                                                    ✔ 
Linux greg-virtualbox 6.9.2-1-MANJARO #1 SMP PREEMPT_DYNAMIC Mon May 27 03:56:18 UTC 2024 x86_64 GNU/Linux
    ~  lsblk                                                                                                                                                       ✔ 
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
sda      8:0    0   40G  0 disk 
├─sda1   8:1    0   40G  0 part /run/BtrfsAssistant/c863c9b5-d87a-4196-bdfb-5002d20e4ce0
└─sda2   8:2    0    8M  0 part 
sr0     11:0    1 1024M  0 rom  
    ~  lsblk -f                                                                                                                                                    ✔ 
NAME   FSTYPE FSVER LABEL UUID                                 FSAVAIL FSUSE% MOUNTPOINTS
sda                                                                           
├─sda1 btrfs              c863c9b5-d87a-4196-bdfb-5002d20e4ce0                /var/log
└─sda2                                                                        
sr0                                                                           
    ~  lsblk -fs                                                                                                                                                   ✔ 
NAME  FSTYPE FSVER LABEL UUID                                 FSAVAIL FSUSE% MOUNTPOINTS
sda1  btrfs              c863c9b5-d87a-4196-bdfb-5002d20e4ce0                /var/log
└─sda                                                                        
sda2                                                                         
└─sda                                                                        
sr0                                                                          
    ~  lsblk -p                                                                                                                                                    ✔ 
NAME        MAJ:MIN RM  SIZE RO TYPE MOUNTPOINTS
/dev/sda      8:0    0   40G  0 disk 
├─/dev/sda1   8:1    0   40G  0 part /var/log
└─/dev/sda2   8:2    0    8M  0 part 
/dev/sr0     11:0    1 1024M  0 rom  
    ~  lsblk -m                                                                                                                                                    ✔ 
NAME    SIZE OWNER GROUP   MODE
sda      40G root  disk    brw-rw----
├─sda1   40G root  disk    brw-rw----
└─sda2    8M root  disk    brw-rw----
sr0    1024M root  optical brw-rw----
    ~                                                       

Sorry about the formatting from zsh, iv never got around to finding out how to copy/paste zsh output and remove the unicode, im on a fresh install of kde on a VB

[Molski]

@jackdinn

Linux greg-virtualbox 6.9.2-1-MANJARO

If you were unsatisfied with the stability of updates and whatnot, why not stick with LTS kernels?

[jackdinn]

this is a virtualbox fresh install. It is where im testing, it has whatever it has after installation from the downloaded iso.

My main system does run the latest LTS but i have not started testing snapper on there just yet (no comfortable enough yet).