For several years, at least 5, when there's a new release I've often done an install into the same btrfs as other installs, having set up those other installs to use subvolumes other than @ and @home. I choose manual install and select the file system where I want it to go, and set it put "/" there. I don't set up /boot/efi anywhere; calamares warns about it but carries on. I set no swap partition.
Now with 25.04 calamares crashes shortly after starting the install with:
@ and @home do not exist before running calamares. (They do after the crash, mounted somewhere under /tmp.)
Is it still possible to install into @ and @home on an existing btrfs? (It would be good not to go back to shuffling partitions around.)
I've had calamares crashes with previous releases, usually involving a swap file, and I've been able to get past them by hacking the Python to just not run the code that crashes. But this crash seems not so simple.
Now with 25.04 calamares crashes shortly after starting the install with:
Code:
Create subvolume '/tmp/calamares-root-nqq5css3/@' Create subvolume '/tmp/calamares-root-nqq5css3/@home' ERROR: Could not create subvolume: File exists Python Error: <class 'subprocess.CalledProcessError'> None File "/usr/lib/x86_64-linux-gnu/calamares/modules/mount/main.py", line 384, in run mount_partition(root_mount_point, partition, partitions, mount_options, mount_options_list, efi_location) ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/lib/x86_64-linux-gnu/calamares/modules/mount/main.py", line 292, in mount_partition subprocess.check_call(["btrfs", "subvolume", "create", ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ root_mount_point + s["subvolume"]]) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Is it still possible to install into @ and @home on an existing btrfs? (It would be good not to go back to shuffling partitions around.)
I've had calamares crashes with previous releases, usually involving a swap file, and I've been able to get past them by hacking the Python to just not run the code that crashes. But this crash seems not so simple.
Comment