/tmp running out of space

DNS stopped resolving on my Turris Omnia and after logging into it over ssh this seems to be because /tmp has run out of space.

Also the web UI is failing with “type ‘exceptions.IOError’” “A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.”

root@turris:~# uptime
19:11:37 up 125 days, 13:07, load average: 0.00, 0.00, 0.00
root@turris:~# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mmcblk0p1 7633920 261204 7375792 3% /
tmpfs 1035384 1035384 0 100% /tmp
tmpfs 512 4 508 1% /dev
root@turris:~# ls -lh /tmp/log
drwxr-xr-x 2 root root 40 Mar 12 05:05 ddns
-rw------- 1 root root 929.6M Jul 15 19:11 iptables
-rw-r–r-- 1 root root 0 Mar 12 05:04 lastlog
drwxr-xr-x 2 root root 60 Mar 12 05:05 lighttpd
-rw------- 1 root root 76.8M Jul 15 19:11 messages
-rw-r–r-- 1 root root 434 Mar 12 05:04 ulogd.log
-rw-r–r-- 1 root root 24 Mar 12 05:04 ulogd.pcap
-rw-r–r-- 1 root root 0 Mar 12 05:04 wtmp

Looks like there is no logrotation configured by default?

It looks like for some reason automatic updates were stuck?
I first tried just removing the iptables log and restarting the ng-syslogd, but it was still misbehaving.
Now after a reboot it seems to have updated to a newer kernel as well:
Linux turris 4.4.74-967673b9d511e4292e3bcb76c9e064bc-0 #1 SMP Thu Jun 29 23:02:30 CEST 2017 armv7l n
I didn’t look at the kernel version before the reboot, but given the 125 day uptime it must have been much older.

I hit the same issue:
root@turris:/tmp/log# ls -lh
-rw------- 1 root root 459.4M Sep 14 20:15 conntrackd-stats.log

any way to make sure that it will not happen again?