cc_resizefs does not allocate space on all disks of a multi disk btrfs volume

[Richard Schwab]

cloud provider: self-hosted, proxmox

cloud-init config:
growpart:
  mode: auto
  devices:
    - /dev/sda1
    - /dev/sdb1
    - /dev/sdc1
    - /dev/sdd1

I have a multi disk btrfs volume as root fs for the vm.
Each disk needs to grow on first boot and the btrfs filesystem should be resized.
Partitions are grown just fine and btrfs detects that, however, running `btrfs filesystem resize max /` as currently used in https://github.com/canonical/cloud-init/blob/bae9b11da9ed7dd0b16fe5adeaf4774b7cc628cf/cloudinit/config/cc_resizefs.py#L81 does not account for having multiple disks.

From the `btrfs-filesystem` man page (debian 11, Btrfs v5.10.1):
> The devid can be found in the output of btrfs filesystem show and defaults to 1 if not specified.

Since a devid is not passed only the first disk will be resized, other disks will stay with "slack" space.

This is after I manually resized disk id 2 for testing, before it also had 3GiB slack:

root@test-vm:~# btrfs device usage /
/dev/sda1, ID: 1
   Device size: 4.00GiB
   Device slack: 3.50KiB
   Data,RAID10: 310.25MiB
   Metadata,RAID10: 64.00MiB
   System,RAID10: 8.00MiB
   Unallocated: 3.62GiB

/dev/sdb1, ID: 2
   Device size: 4.00GiB
   Device slack: 3.50KiB
   Data,RAID10: 310.25MiB
   Metadata,RAID10: 64.00MiB
   System,RAID10: 8.00MiB
   Unallocated: 3.62GiB

/dev/sdc1, ID: 3
   Device size: 4.00GiB
   Device slack: 3.00GiB
   Data,RAID10: 310.25MiB
   Metadata,RAID10: 64.00MiB
   System,RAID10: 8.00MiB
   Unallocated: 640.73MiB

/dev/sdd1, ID: 4
   Device size: 4.00GiB
   Device slack: 3.00GiB
   Data,RAID10: 310.25MiB
   Metadata,RAID10: 64.00MiB
   System,RAID10: 8.00MiB
   Unallocated: 640.73MiB

[James Falcon (falcojr)]

Thank you for the detailed report. This is a use case we should support.

[James Falcon (falcojr)]

Tracked in Github Issues as https://github.com/canonical/cloud-init/issues/3937