HostDiskSpace alert doesn't trigger with mountpoints below /

[guimaluf]

Bug Description

If we have a mountpoint for anything under / the alert doesn't trigger when the disk usage of this mountpoint is above the threshold.

That seems to be due the HostDiskSpace metric uses a regex matching plain / and not everything within it:
used_disk_space{mountpoint=~"/"} > 90

A lazy approach would be changing to mountpoint=~"/.*". That way, everything below / would match the alert.

To Reproduce

On any monitored system

# create a dummy file
fallocate -l10G /var/tmp/host_disk_space
# get a free loop device
losetup -f 
# use the free loop device
losetup /dev/loop9 /var/tmp/host_disk_space
# create the file system
mkfs -t ext4 /dev/loop9
# mount
mount /dev/loop9 /mnt/
# fill up the mountpoint
fallocate -l9G /mnt/fillup

No alerts are generated that way.

Environment

n/a

Relevant log output

n/a

Additional context

No response

[sed-i]

Thanks @guimaluf!
We should also fix this in the VM charm.