Closed vajonam closed 3 years ago
$df -m /
Filesystem 1M-blocks Used Avail Capacity Mounted on
/dev/gptid/6f34ba9a-3faa-11ea-bfde-40623108486d 13683 3328 9260 26% /
This shows 3328
megs used
and when running DU. this shows differences!
$cd / && du -ma | sort -nr | head -n 20
2124 .
1227 ./usr
851 ./usr/local
605 ./var
512 ./var/unbound
474 ./var/unbound/usr/local
474 ./var/unbound/usr
316 ./var/unbound/usr/local/lib
316 ./usr/local/lib
298 ./usr/lib
249 ./usr/local/share
207 ./boot
132 ./boot/kernel.old
130 ./var/unbound/usr/local/lib/python3.7
130 ./usr/local/lib/python3.7
115 ./var/unbound/usr/local/bin
115 ./usr/local/bin
113 ./usr/lib/debug
102 ./usr/local/sbin
82 ./var/db
As you can see du reports 2124
megs used and df reports 3328
Think this is due the telegraf_unbound.sh .. have moved over to the lite stuff now. will provide an update shortly.
Yup definitely due to the telegraph_unbound.sh, not sure if this is related to python mode and the pfblocker. but switching to unbound_lite seemed to fix this issue.
I am seeing an issue that is being discussed here.
https://www.reddit.com/r/pfBlockerNG/comments/md1kli/pfblocker_using_up_disk_space_over_a_few_days/
Originally I thought it was to do w/ pfblocker but it seems to do with thee telegraf log tail/parser that is keeping the files open and not let it be deleted. What ends up happening is a difference between
df
anddu
commands. restart pfblocker everything resets. Restarting telegraf doesn't seem to help.this is what lsof shows.
removing the telegraph log tail from the additional config seems to fix it.
Are you running into this? this is with pfblocker_devel (latest) and pfsense 2.5.0. Any ideas what I can do here?