To small root-partition. Reinstallastion?

Hi!
Have an old Dell Studio17 with an 110GB SSD. Disk is setup with 40GB root and 70GB /home.
Root is BtrFS and /Home is /ext4

The problem is that root-partition is full and i can’t do zypper dup anymore. Everytime i need to delete Snapper snapshots to free diskspace.
So i want to reduce /home and increase /.
But i’m thinking to maybe reinstall the whole system. This i can delete junk that i don’t need anymore.

I it possible to do an reinstall and reduce /home without loosing /home?

I don’t use BRTFS myself, but I note from this document:
https://doc.opensuse.org/documentation/leap/reference/html/book.opensuse.reference/cha.expert_partitioner.html

the suggested size for the root partition is:

“*The suggested size for a root Btrfs partition with default subvolumes is 20 GB. *”

Perhaps you have other “stuff” in the root partition, maybe /tmp and /var/tmp are not being regularly purged.

With regard to resizing an existing partition, generally yes, but the partition can’t be mounted, so you would need to carry that out from a live cd/usb. I’d ensure you have a reliable backup of your home partition data first though.

Hi
Have you configured snapper to reduce the amount of snapshots kept, have the btrfs maintenance routines been run/are running properly?

If you decide you really want a larger root partition, the ext4 partition can be resized / moved with a partition manager and then the root partition can be grown.

If you decide to reinstall, my experience so far with the newer Tumbleweed installs made me feel comfortable with using a full disk installation using btrfs. There is a recent discussion you can search for in the forum that has the various advantages/disadvantages of having a separate home partition. Before using Tumbleweed/btrfs I used a separate home partition to be safe when the quarterly or semiannual upgrades took place. With Tumbleweed/btrfs there hasn’t been a need for a full reinstall.

Maybe that works for Leap. But it is too small for Tumbleweed.

I currently have Leap 15.2 Alpha (on a test system) with a 50G “btrfs” root file system. And when I last checked, it was 71% full. Some of that will probably be freed up before the next major update. At present, in its alpha testing stage, Leap 15.2 is in effect a rolling release much like Tumbleweed. Those snapshots fill up quickly with all of the updates.

I would suggest at least 50G for Tumbleweed “btrfs” partition. And may even 80G might be a good idea.

Like you, I mostly avoid “btrfs”, but I occasionally use it in test systems. My current desktop is all “ext4”.

Hi
I don’t use snapper on tumbleweed, if push comes to shove and the system comes up can use tumbleweed-cli to rolback. Rest is xfs which have had no issues with. $HOME for me is on btrfs.

Or the documentation is rather outdated… :\

Apart from discussing all sorts of docymentation advices and/or installation defualts, I think that 40GB should be enough for a normal Btrfs installation. So shouldn’t we try to find out why it is full, Or begin asking the OP why he thinks it is full (so far we only have the OP’s conclusion, no computerfacts).

And a usefull suggestion is malcolm’s to check if there aren’t too many snapshots.

And to the OP, apart from that check on snapshots, please show

df -h

If there is anything you absolutely don’t want to risk, backup.

Then,
Download the following and depending on whether your system supports a CD/DVD and/or USB bootloader,
Burn/copy to the device

https://gparted.org/livecd.php

Boot to your GParted Live,
Modify your partitions as you like and
Reboot.
Done, problem fixed.

Regardless of recommendations, my personal recommendation is that a BTRFS partition with snapshots enabled should be minimum 100GB.
Or, someone should re-write the default snapper configuration to more aggressively remove old snapshots automatically, and do it based on a free disk space threshold. Last I looked at the Snapper configuration, I think some efforts have been made in this direction but more needs to be done.

TSU

A reminder that the common “df” command reports inaccurately on BTRFS…
You have to prepend “btrfs filesystem” to the command to return accurate results.
(and the word “filesystem” can be abbreviated with simply “fi”)

The SDB:BTRFS contains some good examples how to run useful BTRFS filesystem commands including the following

Display common properties for the root partition

## btrfs filesystem show *mount_point *
btrfs filesystem show /

The “df” command on the root partition

## btrfs filesystem df *mount_point*
btrfs filesystem df /

The “usage” command on the root partition

## btrfs filesystem usage *mount_point*
btrfs filesystem usage /

BTW - The SDB also includes a recommended procedure for when your partition is full due to excessive number of snapshots.

TSU