This is a minor issue, but I'm curious about it.
I have a couple of disk space usage tests on /home (debian linux, ext4).
'df -h' reports:
/dev/nvme1n1p1 5.9G 3.7G 2.2G 64% /home
I get the same percentage output regardless of '-h' (powers of 1024) or
'-H' (powers of 1000). The gigabyte values differ, the percentage remains
the same.
Tune2fs reports no reserved blocks on the partition:
tune2fs -l /dev/nvme1n1p1 | grep 'Reserved'
Reserved block count: 0
Reserved GDT blocks: 511
Reserved blocks uid: 0 (user root)
Reserved blocks gid: 0 (group root)
But Monit's status output reports:
space total 5.8 GB (of which 0.3% is reserved for root user)"
my disk space argument is:
# HOME
check filesystem homefs with path /home
every 5 cycles
if space greater 63 %
then exec /usr/local/bin/pushover
repeat every 5 cycles
The problem is, I will begin getting alerts before it reached 64% - as
reported by df - example:
host [xxxxx] homefs - space usage 63.0% matches
resource limit [space usage > 63.0%
As I said, minor - very minor - issue, though it is curious that Monit
thinks there are reserved blocks on a partition with none. It'd just be
great if there were a way to conform the output to match df's output.
--
Paul Theodoropoulos
www.anastrophe.com
