It doesn't work with the increased btrfs partition size because of the wrong padding between partitions (the thing I was mentioning on MM). I get output similar to this from parted (it's in the cloud-init logs):
Warning: You requested a partition from 10.0GB to 107GB (sectors
19531250..209715199).
The closest location we can manage is 10.6GB to 107GB (sectors
20686848..209715166).
and then the partition is *not* created. By manually trying I found out that a 600MB padding works with a 10GB root partition:
It doesn't work with the increased btrfs partition size because of the wrong padding between partitions (the thing I was mentioning on MM). I get output similar to this from parted (it's in the cloud-init logs):
Warning: You requested a partition from 10.0GB to 107GB (sectors .209715199) . .209715166) .
19531250.
The closest location we can manage is 10.6GB to 107GB (sectors
20686848.
and then the partition is *not* created. By manually trying I found out that a 600MB padding works with a 10GB root partition:
- echo 'mkpart primary btrfs 10600MB 100%\n' | parted /dev/sda
This is kind of ugly and arbitrary, the difference may be due to one tool using MB and the other MiB.